{ "best_metric": null, "best_model_checkpoint": null, "epoch": 20.0, "eval_steps": 500, "global_step": 15620, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.012804097311139564, "grad_norm": 95.68589782714844, "learning_rate": 6.402048655569782e-08, "loss": 1.1693, "step": 10 }, { "epoch": 0.02560819462227913, "grad_norm": 83.14017486572266, "learning_rate": 1.2804097311139564e-07, "loss": 1.4169, "step": 20 }, { "epoch": 0.03841229193341869, "grad_norm": 63.35808563232422, "learning_rate": 1.9206145966709346e-07, "loss": 1.0669, "step": 30 }, { "epoch": 0.05121638924455826, "grad_norm": 2.2753270968678407e-05, "learning_rate": 2.560819462227913e-07, "loss": 0.7246, "step": 40 }, { "epoch": 0.06402048655569782, "grad_norm": 83.9839859008789, "learning_rate": 3.2010243277848913e-07, "loss": 1.9526, "step": 50 }, { "epoch": 0.07682458386683738, "grad_norm": 127.40715789794922, "learning_rate": 3.841229193341869e-07, "loss": 1.2232, "step": 60 }, { "epoch": 0.08962868117797695, "grad_norm": 58.6407470703125, "learning_rate": 4.481434058898848e-07, "loss": 0.6957, "step": 70 }, { "epoch": 0.10243277848911651, "grad_norm": 42.71138000488281, "learning_rate": 5.121638924455826e-07, "loss": 0.7531, "step": 80 }, { "epoch": 0.11523687580025609, "grad_norm": 75.90082550048828, "learning_rate": 5.761843790012805e-07, "loss": 1.3796, "step": 90 }, { "epoch": 0.12804097311139565, "grad_norm": 53.02857208251953, "learning_rate": 6.402048655569783e-07, "loss": 0.4958, "step": 100 }, { "epoch": 0.14084507042253522, "grad_norm": 0.12084757536649704, "learning_rate": 7.042253521126762e-07, "loss": 1.2211, "step": 110 }, { "epoch": 0.15364916773367476, "grad_norm": 101.07658386230469, "learning_rate": 7.682458386683739e-07, "loss": 0.7834, "step": 120 }, { "epoch": 0.16645326504481434, "grad_norm": 42.64153289794922, "learning_rate": 8.322663252240718e-07, "loss": 1.3529, "step": 130 }, { "epoch": 0.1792573623559539, "grad_norm": 79.7669906616211, "learning_rate": 8.962868117797697e-07, "loss": 0.8512, "step": 140 }, { "epoch": 0.19206145966709348, "grad_norm": 73.68550109863281, "learning_rate": 9.603072983354676e-07, "loss": 0.5557, "step": 150 }, { "epoch": 0.20486555697823303, "grad_norm": 89.7555923461914, "learning_rate": 1.0243277848911651e-06, "loss": 1.0599, "step": 160 }, { "epoch": 0.2176696542893726, "grad_norm": 65.42781829833984, "learning_rate": 1.0883482714468631e-06, "loss": 1.1241, "step": 170 }, { "epoch": 0.23047375160051217, "grad_norm": 81.75363159179688, "learning_rate": 1.152368758002561e-06, "loss": 1.5703, "step": 180 }, { "epoch": 0.24327784891165172, "grad_norm": 40.48162841796875, "learning_rate": 1.2163892445582587e-06, "loss": 0.8459, "step": 190 }, { "epoch": 0.2560819462227913, "grad_norm": 49.82510757446289, "learning_rate": 1.2804097311139565e-06, "loss": 0.6623, "step": 200 }, { "epoch": 0.26888604353393086, "grad_norm": 47.12430953979492, "learning_rate": 1.3444302176696545e-06, "loss": 0.7989, "step": 210 }, { "epoch": 0.28169014084507044, "grad_norm": 79.40914916992188, "learning_rate": 1.4084507042253523e-06, "loss": 0.8171, "step": 220 }, { "epoch": 0.29449423815621, "grad_norm": 26.3239803314209, "learning_rate": 1.4724711907810501e-06, "loss": 1.2716, "step": 230 }, { "epoch": 0.3072983354673495, "grad_norm": 67.35940551757812, "learning_rate": 1.5364916773367477e-06, "loss": 0.8925, "step": 240 }, { "epoch": 0.3201024327784891, "grad_norm": 53.327667236328125, "learning_rate": 1.6005121638924457e-06, "loss": 0.7207, "step": 250 }, { "epoch": 0.33290653008962867, "grad_norm": 7.487966691144265e-09, "learning_rate": 1.6645326504481435e-06, "loss": 0.6223, "step": 260 }, { "epoch": 0.34571062740076824, "grad_norm": 5.723132133483887, "learning_rate": 1.7285531370038413e-06, "loss": 0.6425, "step": 270 }, { "epoch": 0.3585147247119078, "grad_norm": 42.45857620239258, "learning_rate": 1.7925736235595393e-06, "loss": 0.635, "step": 280 }, { "epoch": 0.3713188220230474, "grad_norm": 59.34583282470703, "learning_rate": 1.856594110115237e-06, "loss": 0.6766, "step": 290 }, { "epoch": 0.38412291933418696, "grad_norm": 4.818227807845687e-07, "learning_rate": 1.920614596670935e-06, "loss": 0.5873, "step": 300 }, { "epoch": 0.3969270166453265, "grad_norm": 31.084566116333008, "learning_rate": 1.9846350832266325e-06, "loss": 0.3204, "step": 310 }, { "epoch": 0.40973111395646605, "grad_norm": 33.57612991333008, "learning_rate": 2.0486555697823303e-06, "loss": 0.5008, "step": 320 }, { "epoch": 0.4225352112676056, "grad_norm": 107.3833999633789, "learning_rate": 2.1126760563380285e-06, "loss": 0.5925, "step": 330 }, { "epoch": 0.4353393085787452, "grad_norm": 82.26516723632812, "learning_rate": 2.1766965428937263e-06, "loss": 0.6275, "step": 340 }, { "epoch": 0.44814340588988477, "grad_norm": 37.221885681152344, "learning_rate": 2.240717029449424e-06, "loss": 0.8503, "step": 350 }, { "epoch": 0.46094750320102434, "grad_norm": 55.0982780456543, "learning_rate": 2.304737516005122e-06, "loss": 0.2618, "step": 360 }, { "epoch": 0.4737516005121639, "grad_norm": 4.648223139724905e-08, "learning_rate": 2.3687580025608197e-06, "loss": 0.4732, "step": 370 }, { "epoch": 0.48655569782330343, "grad_norm": 71.33980560302734, "learning_rate": 2.4327784891165175e-06, "loss": 0.5351, "step": 380 }, { "epoch": 0.499359795134443, "grad_norm": 58.13515090942383, "learning_rate": 2.4967989756722153e-06, "loss": 0.972, "step": 390 }, { "epoch": 0.5121638924455826, "grad_norm": 50.091941833496094, "learning_rate": 2.560819462227913e-06, "loss": 0.6998, "step": 400 }, { "epoch": 0.5249679897567221, "grad_norm": 53.93059539794922, "learning_rate": 2.624839948783611e-06, "loss": 0.501, "step": 410 }, { "epoch": 0.5377720870678617, "grad_norm": 54.664817810058594, "learning_rate": 2.688860435339309e-06, "loss": 0.6961, "step": 420 }, { "epoch": 0.5505761843790012, "grad_norm": 37.57455062866211, "learning_rate": 2.7528809218950064e-06, "loss": 0.2497, "step": 430 }, { "epoch": 0.5633802816901409, "grad_norm": 37.04225540161133, "learning_rate": 2.8169014084507046e-06, "loss": 0.4855, "step": 440 }, { "epoch": 0.5761843790012804, "grad_norm": 55.897037506103516, "learning_rate": 2.880921895006402e-06, "loss": 0.1553, "step": 450 }, { "epoch": 0.58898847631242, "grad_norm": 30.655546188354492, "learning_rate": 2.9449423815621002e-06, "loss": 0.6863, "step": 460 }, { "epoch": 0.6017925736235595, "grad_norm": 54.875457763671875, "learning_rate": 3.008962868117798e-06, "loss": 0.5614, "step": 470 }, { "epoch": 0.614596670934699, "grad_norm": 0.04425310343503952, "learning_rate": 3.0729833546734954e-06, "loss": 0.1848, "step": 480 }, { "epoch": 0.6274007682458387, "grad_norm": 24.247211456298828, "learning_rate": 3.1370038412291936e-06, "loss": 0.4968, "step": 490 }, { "epoch": 0.6402048655569782, "grad_norm": 36.39949417114258, "learning_rate": 3.2010243277848914e-06, "loss": 0.4316, "step": 500 }, { "epoch": 0.6530089628681178, "grad_norm": 30.007171630859375, "learning_rate": 3.265044814340589e-06, "loss": 0.3079, "step": 510 }, { "epoch": 0.6658130601792573, "grad_norm": 93.54869842529297, "learning_rate": 3.329065300896287e-06, "loss": 0.7278, "step": 520 }, { "epoch": 0.678617157490397, "grad_norm": 33.79115676879883, "learning_rate": 3.3930857874519852e-06, "loss": 0.4887, "step": 530 }, { "epoch": 0.6914212548015365, "grad_norm": 29.5987606048584, "learning_rate": 3.4571062740076826e-06, "loss": 0.856, "step": 540 }, { "epoch": 0.704225352112676, "grad_norm": 50.13193893432617, "learning_rate": 3.5211267605633804e-06, "loss": 0.4866, "step": 550 }, { "epoch": 0.7170294494238156, "grad_norm": 33.70960235595703, "learning_rate": 3.5851472471190786e-06, "loss": 0.5089, "step": 560 }, { "epoch": 0.7298335467349552, "grad_norm": 0.006306682713329792, "learning_rate": 3.649167733674776e-06, "loss": 0.3933, "step": 570 }, { "epoch": 0.7426376440460948, "grad_norm": 41.15188980102539, "learning_rate": 3.713188220230474e-06, "loss": 0.3824, "step": 580 }, { "epoch": 0.7554417413572343, "grad_norm": 31.334068298339844, "learning_rate": 3.777208706786172e-06, "loss": 0.4835, "step": 590 }, { "epoch": 0.7682458386683739, "grad_norm": 36.58442306518555, "learning_rate": 3.84122919334187e-06, "loss": 0.6963, "step": 600 }, { "epoch": 0.7810499359795134, "grad_norm": 28.94176483154297, "learning_rate": 3.9052496798975676e-06, "loss": 0.5345, "step": 610 }, { "epoch": 0.793854033290653, "grad_norm": 2.0806937754969113e-05, "learning_rate": 3.969270166453265e-06, "loss": 0.3802, "step": 620 }, { "epoch": 0.8066581306017926, "grad_norm": 47.475311279296875, "learning_rate": 4.033290653008963e-06, "loss": 0.4064, "step": 630 }, { "epoch": 0.8194622279129321, "grad_norm": 29.445634841918945, "learning_rate": 4.0973111395646605e-06, "loss": 0.3496, "step": 640 }, { "epoch": 0.8322663252240717, "grad_norm": 30.454071044921875, "learning_rate": 4.161331626120359e-06, "loss": 0.2923, "step": 650 }, { "epoch": 0.8450704225352113, "grad_norm": 32.31270980834961, "learning_rate": 4.225352112676057e-06, "loss": 0.2513, "step": 660 }, { "epoch": 0.8578745198463509, "grad_norm": 24.27973747253418, "learning_rate": 4.289372599231754e-06, "loss": 0.2114, "step": 670 }, { "epoch": 0.8706786171574904, "grad_norm": 39.10868835449219, "learning_rate": 4.3533930857874526e-06, "loss": 0.5525, "step": 680 }, { "epoch": 0.8834827144686299, "grad_norm": 41.959476470947266, "learning_rate": 4.41741357234315e-06, "loss": 0.6292, "step": 690 }, { "epoch": 0.8962868117797695, "grad_norm": 82.65428924560547, "learning_rate": 4.481434058898848e-06, "loss": 0.4392, "step": 700 }, { "epoch": 0.9090909090909091, "grad_norm": 30.887210845947266, "learning_rate": 4.5454545454545455e-06, "loss": 0.2588, "step": 710 }, { "epoch": 0.9218950064020487, "grad_norm": 30.215349197387695, "learning_rate": 4.609475032010244e-06, "loss": 0.2345, "step": 720 }, { "epoch": 0.9346991037131882, "grad_norm": 18.30318260192871, "learning_rate": 4.673495518565941e-06, "loss": 0.1388, "step": 730 }, { "epoch": 0.9475032010243278, "grad_norm": 30.345367431640625, "learning_rate": 4.737516005121639e-06, "loss": 0.2127, "step": 740 }, { "epoch": 0.9603072983354674, "grad_norm": 35.18952178955078, "learning_rate": 4.8015364916773375e-06, "loss": 0.3856, "step": 750 }, { "epoch": 0.9731113956466069, "grad_norm": 52.910892486572266, "learning_rate": 4.865556978233035e-06, "loss": 0.6051, "step": 760 }, { "epoch": 0.9859154929577465, "grad_norm": 32.33030319213867, "learning_rate": 4.929577464788733e-06, "loss": 0.3884, "step": 770 }, { "epoch": 0.998719590268886, "grad_norm": 1.3143353072564423e-09, "learning_rate": 4.9935979513444305e-06, "loss": 0.6735, "step": 780 }, { "epoch": 1.0115236875800255, "grad_norm": 5.95734167098999, "learning_rate": 5.057618437900128e-06, "loss": 0.1643, "step": 790 }, { "epoch": 1.0243277848911652, "grad_norm": 32.84043884277344, "learning_rate": 5.121638924455826e-06, "loss": 0.2823, "step": 800 }, { "epoch": 1.0371318822023048, "grad_norm": 9.927327156066895, "learning_rate": 5.185659411011524e-06, "loss": 0.3924, "step": 810 }, { "epoch": 1.0499359795134442, "grad_norm": 178.80641174316406, "learning_rate": 5.249679897567222e-06, "loss": 0.3735, "step": 820 }, { "epoch": 1.0627400768245838, "grad_norm": 24.103487014770508, "learning_rate": 5.31370038412292e-06, "loss": 0.2474, "step": 830 }, { "epoch": 1.0755441741357235, "grad_norm": 37.77121353149414, "learning_rate": 5.377720870678618e-06, "loss": 0.219, "step": 840 }, { "epoch": 1.088348271446863, "grad_norm": 2.87650728225708, "learning_rate": 5.4417413572343155e-06, "loss": 0.2882, "step": 850 }, { "epoch": 1.1011523687580025, "grad_norm": 43.957557678222656, "learning_rate": 5.505761843790013e-06, "loss": 0.4412, "step": 860 }, { "epoch": 1.1139564660691421, "grad_norm": 45.359195709228516, "learning_rate": 5.569782330345711e-06, "loss": 0.4331, "step": 870 }, { "epoch": 1.1267605633802817, "grad_norm": 23.766220092773438, "learning_rate": 5.633802816901409e-06, "loss": 0.2998, "step": 880 }, { "epoch": 1.1395646606914211, "grad_norm": 41.96795654296875, "learning_rate": 5.697823303457106e-06, "loss": 0.4659, "step": 890 }, { "epoch": 1.1523687580025608, "grad_norm": 37.043582916259766, "learning_rate": 5.761843790012804e-06, "loss": 0.4146, "step": 900 }, { "epoch": 1.1651728553137004, "grad_norm": 37.70871353149414, "learning_rate": 5.825864276568502e-06, "loss": 0.2674, "step": 910 }, { "epoch": 1.17797695262484, "grad_norm": 0.0048003848642110825, "learning_rate": 5.8898847631242005e-06, "loss": 0.1727, "step": 920 }, { "epoch": 1.1907810499359794, "grad_norm": 0.0003311179461888969, "learning_rate": 5.953905249679898e-06, "loss": 0.1689, "step": 930 }, { "epoch": 1.203585147247119, "grad_norm": 19.235780715942383, "learning_rate": 6.017925736235596e-06, "loss": 0.2137, "step": 940 }, { "epoch": 1.2163892445582587, "grad_norm": 0.0003354971413500607, "learning_rate": 6.081946222791294e-06, "loss": 0.243, "step": 950 }, { "epoch": 1.229193341869398, "grad_norm": 26.026771545410156, "learning_rate": 6.145966709346991e-06, "loss": 0.1814, "step": 960 }, { "epoch": 1.2419974391805377, "grad_norm": 39.604984283447266, "learning_rate": 6.209987195902689e-06, "loss": 0.2839, "step": 970 }, { "epoch": 1.2548015364916774, "grad_norm": 18.712791442871094, "learning_rate": 6.274007682458387e-06, "loss": 0.1458, "step": 980 }, { "epoch": 1.267605633802817, "grad_norm": 6.844394683837891, "learning_rate": 6.3380281690140855e-06, "loss": 0.2115, "step": 990 }, { "epoch": 1.2804097311139564, "grad_norm": 21.862825393676758, "learning_rate": 6.402048655569783e-06, "loss": 0.3372, "step": 1000 }, { "epoch": 1.293213828425096, "grad_norm": 36.02387619018555, "learning_rate": 6.466069142125481e-06, "loss": 0.283, "step": 1010 }, { "epoch": 1.3060179257362357, "grad_norm": 23.573848724365234, "learning_rate": 6.530089628681178e-06, "loss": 0.2957, "step": 1020 }, { "epoch": 1.318822023047375, "grad_norm": 45.303897857666016, "learning_rate": 6.594110115236876e-06, "loss": 0.2657, "step": 1030 }, { "epoch": 1.3316261203585147, "grad_norm": 32.70471954345703, "learning_rate": 6.658130601792574e-06, "loss": 0.2656, "step": 1040 }, { "epoch": 1.3444302176696543, "grad_norm": 24.811328887939453, "learning_rate": 6.722151088348272e-06, "loss": 0.2209, "step": 1050 }, { "epoch": 1.357234314980794, "grad_norm": 27.601123809814453, "learning_rate": 6.7861715749039704e-06, "loss": 0.2363, "step": 1060 }, { "epoch": 1.3700384122919333, "grad_norm": 31.767385482788086, "learning_rate": 6.850192061459667e-06, "loss": 0.3528, "step": 1070 }, { "epoch": 1.382842509603073, "grad_norm": 17.689594268798828, "learning_rate": 6.914212548015365e-06, "loss": 0.2899, "step": 1080 }, { "epoch": 1.3956466069142126, "grad_norm": 2.022170066833496, "learning_rate": 6.978233034571063e-06, "loss": 0.0892, "step": 1090 }, { "epoch": 1.408450704225352, "grad_norm": 20.169776916503906, "learning_rate": 7.042253521126761e-06, "loss": 0.285, "step": 1100 }, { "epoch": 1.4212548015364916, "grad_norm": 23.703601837158203, "learning_rate": 7.106274007682459e-06, "loss": 0.2353, "step": 1110 }, { "epoch": 1.4340588988476313, "grad_norm": 23.385953903198242, "learning_rate": 7.170294494238157e-06, "loss": 0.2653, "step": 1120 }, { "epoch": 1.446862996158771, "grad_norm": 9.762706213223282e-06, "learning_rate": 7.234314980793855e-06, "loss": 0.2132, "step": 1130 }, { "epoch": 1.4596670934699103, "grad_norm": 16.660194396972656, "learning_rate": 7.298335467349552e-06, "loss": 0.2732, "step": 1140 }, { "epoch": 1.47247119078105, "grad_norm": 0.0010289043420925736, "learning_rate": 7.36235595390525e-06, "loss": 0.3515, "step": 1150 }, { "epoch": 1.4852752880921896, "grad_norm": 6.873804092407227, "learning_rate": 7.426376440460948e-06, "loss": 0.1853, "step": 1160 }, { "epoch": 1.498079385403329, "grad_norm": 0.008222809061408043, "learning_rate": 7.490396927016646e-06, "loss": 0.153, "step": 1170 }, { "epoch": 1.5108834827144686, "grad_norm": 36.408592224121094, "learning_rate": 7.554417413572344e-06, "loss": 0.319, "step": 1180 }, { "epoch": 1.5236875800256082, "grad_norm": 36.3063850402832, "learning_rate": 7.618437900128041e-06, "loss": 0.2851, "step": 1190 }, { "epoch": 1.5364916773367479, "grad_norm": 1.4304002434073482e-05, "learning_rate": 7.68245838668374e-06, "loss": 0.1114, "step": 1200 }, { "epoch": 1.5492957746478875, "grad_norm": 2.674076995390351e-07, "learning_rate": 7.746478873239436e-06, "loss": 0.1014, "step": 1210 }, { "epoch": 1.5620998719590269, "grad_norm": 24.77590560913086, "learning_rate": 7.810499359795135e-06, "loss": 0.3714, "step": 1220 }, { "epoch": 1.5749039692701663, "grad_norm": 21.162446975708008, "learning_rate": 7.874519846350833e-06, "loss": 0.2549, "step": 1230 }, { "epoch": 1.587708066581306, "grad_norm": 9.463348388671875, "learning_rate": 7.93854033290653e-06, "loss": 0.1921, "step": 1240 }, { "epoch": 1.6005121638924455, "grad_norm": 0.0937286987900734, "learning_rate": 8.002560819462229e-06, "loss": 0.1471, "step": 1250 }, { "epoch": 1.6133162612035852, "grad_norm": 8.174616813659668, "learning_rate": 8.066581306017926e-06, "loss": 0.2053, "step": 1260 }, { "epoch": 1.6261203585147248, "grad_norm": 24.255199432373047, "learning_rate": 8.130601792573625e-06, "loss": 0.1601, "step": 1270 }, { "epoch": 1.6389244558258644, "grad_norm": 26.646930694580078, "learning_rate": 8.194622279129321e-06, "loss": 0.2056, "step": 1280 }, { "epoch": 1.6517285531370038, "grad_norm": 1.1445684432983398, "learning_rate": 8.25864276568502e-06, "loss": 0.1828, "step": 1290 }, { "epoch": 1.6645326504481432, "grad_norm": 15.222755432128906, "learning_rate": 8.322663252240718e-06, "loss": 0.1288, "step": 1300 }, { "epoch": 1.6773367477592829, "grad_norm": 24.017253875732422, "learning_rate": 8.386683738796415e-06, "loss": 0.3157, "step": 1310 }, { "epoch": 1.6901408450704225, "grad_norm": 19.139873504638672, "learning_rate": 8.450704225352114e-06, "loss": 0.1815, "step": 1320 }, { "epoch": 1.7029449423815621, "grad_norm": 21.9299373626709, "learning_rate": 8.514724711907811e-06, "loss": 0.2274, "step": 1330 }, { "epoch": 1.7157490396927018, "grad_norm": 0.8286514282226562, "learning_rate": 8.578745198463509e-06, "loss": 0.3246, "step": 1340 }, { "epoch": 1.7285531370038414, "grad_norm": 10.219658851623535, "learning_rate": 8.642765685019206e-06, "loss": 0.242, "step": 1350 }, { "epoch": 1.7413572343149808, "grad_norm": 10.624405860900879, "learning_rate": 8.706786171574905e-06, "loss": 0.0443, "step": 1360 }, { "epoch": 1.7541613316261202, "grad_norm": 11.301095008850098, "learning_rate": 8.770806658130602e-06, "loss": 0.2199, "step": 1370 }, { "epoch": 1.7669654289372598, "grad_norm": 15.419243812561035, "learning_rate": 8.8348271446863e-06, "loss": 0.1234, "step": 1380 }, { "epoch": 1.7797695262483995, "grad_norm": 0.6660290956497192, "learning_rate": 8.898847631241997e-06, "loss": 0.2716, "step": 1390 }, { "epoch": 1.792573623559539, "grad_norm": 5.008613586425781, "learning_rate": 8.962868117797696e-06, "loss": 0.1467, "step": 1400 }, { "epoch": 1.8053777208706787, "grad_norm": 12.315166473388672, "learning_rate": 9.026888604353394e-06, "loss": 0.1521, "step": 1410 }, { "epoch": 1.8181818181818183, "grad_norm": 24.88693618774414, "learning_rate": 9.090909090909091e-06, "loss": 0.1602, "step": 1420 }, { "epoch": 1.8309859154929577, "grad_norm": 22.850038528442383, "learning_rate": 9.15492957746479e-06, "loss": 0.1432, "step": 1430 }, { "epoch": 1.8437900128040972, "grad_norm": 9.984404563903809, "learning_rate": 9.218950064020487e-06, "loss": 0.1976, "step": 1440 }, { "epoch": 1.8565941101152368, "grad_norm": 5.21957540512085, "learning_rate": 9.282970550576185e-06, "loss": 0.2355, "step": 1450 }, { "epoch": 1.8693982074263764, "grad_norm": 14.592249870300293, "learning_rate": 9.346991037131882e-06, "loss": 0.1723, "step": 1460 }, { "epoch": 1.882202304737516, "grad_norm": 11.301413536071777, "learning_rate": 9.411011523687581e-06, "loss": 0.1269, "step": 1470 }, { "epoch": 1.8950064020486557, "grad_norm": 1.9392039775848389, "learning_rate": 9.475032010243279e-06, "loss": 0.2175, "step": 1480 }, { "epoch": 1.9078104993597953, "grad_norm": 5.65443229675293, "learning_rate": 9.539052496798976e-06, "loss": 0.1728, "step": 1490 }, { "epoch": 1.9206145966709347, "grad_norm": 18.690357208251953, "learning_rate": 9.603072983354675e-06, "loss": 0.1626, "step": 1500 }, { "epoch": 1.933418693982074, "grad_norm": 0.4743829071521759, "learning_rate": 9.667093469910372e-06, "loss": 0.0567, "step": 1510 }, { "epoch": 1.9462227912932137, "grad_norm": 0.1901559680700302, "learning_rate": 9.73111395646607e-06, "loss": 0.0758, "step": 1520 }, { "epoch": 1.9590268886043534, "grad_norm": 10.892840385437012, "learning_rate": 9.795134443021767e-06, "loss": 0.107, "step": 1530 }, { "epoch": 1.971830985915493, "grad_norm": 12.538705825805664, "learning_rate": 9.859154929577466e-06, "loss": 0.1921, "step": 1540 }, { "epoch": 1.9846350832266326, "grad_norm": 11.1087064743042, "learning_rate": 9.923175416133164e-06, "loss": 0.1633, "step": 1550 }, { "epoch": 1.9974391805377723, "grad_norm": 10.895771026611328, "learning_rate": 9.987195902688861e-06, "loss": 0.2034, "step": 1560 }, { "epoch": 2.0102432778489114, "grad_norm": 0.03142176568508148, "learning_rate": 9.994309290083939e-06, "loss": 0.1048, "step": 1570 }, { "epoch": 2.023047375160051, "grad_norm": 0.010399126447737217, "learning_rate": 9.987195902688861e-06, "loss": 0.022, "step": 1580 }, { "epoch": 2.0358514724711907, "grad_norm": 0.01455871295183897, "learning_rate": 9.980082515293785e-06, "loss": 0.2355, "step": 1590 }, { "epoch": 2.0486555697823303, "grad_norm": 9.397605895996094, "learning_rate": 9.972969127898707e-06, "loss": 0.168, "step": 1600 }, { "epoch": 2.06145966709347, "grad_norm": 14.55268669128418, "learning_rate": 9.965855740503629e-06, "loss": 0.1218, "step": 1610 }, { "epoch": 2.0742637644046096, "grad_norm": 7.501183986663818, "learning_rate": 9.958742353108552e-06, "loss": 0.1399, "step": 1620 }, { "epoch": 2.087067861715749, "grad_norm": 0.07658198475837708, "learning_rate": 9.951628965713474e-06, "loss": 0.083, "step": 1630 }, { "epoch": 2.0998719590268884, "grad_norm": 12.069207191467285, "learning_rate": 9.944515578318396e-06, "loss": 0.1421, "step": 1640 }, { "epoch": 2.112676056338028, "grad_norm": 9.84998607635498, "learning_rate": 9.937402190923318e-06, "loss": 0.099, "step": 1650 }, { "epoch": 2.1254801536491676, "grad_norm": 9.7352876663208, "learning_rate": 9.93028880352824e-06, "loss": 0.0748, "step": 1660 }, { "epoch": 2.1382842509603073, "grad_norm": 11.115337371826172, "learning_rate": 9.923175416133164e-06, "loss": 0.1116, "step": 1670 }, { "epoch": 2.151088348271447, "grad_norm": 14.003682136535645, "learning_rate": 9.916062028738086e-06, "loss": 0.0845, "step": 1680 }, { "epoch": 2.1638924455825865, "grad_norm": 8.884621620178223, "learning_rate": 9.908948641343008e-06, "loss": 0.08, "step": 1690 }, { "epoch": 2.176696542893726, "grad_norm": 15.43350887298584, "learning_rate": 9.901835253947931e-06, "loss": 0.0901, "step": 1700 }, { "epoch": 2.189500640204866, "grad_norm": 1.7564079761505127, "learning_rate": 9.894721866552853e-06, "loss": 0.1159, "step": 1710 }, { "epoch": 2.202304737516005, "grad_norm": 10.379612922668457, "learning_rate": 9.887608479157775e-06, "loss": 0.073, "step": 1720 }, { "epoch": 2.2151088348271446, "grad_norm": 8.157423973083496, "learning_rate": 9.880495091762699e-06, "loss": 0.0896, "step": 1730 }, { "epoch": 2.2279129321382842, "grad_norm": 0.7226857542991638, "learning_rate": 9.87338170436762e-06, "loss": 0.1031, "step": 1740 }, { "epoch": 2.240717029449424, "grad_norm": 6.456130027770996, "learning_rate": 9.866268316972543e-06, "loss": 0.0389, "step": 1750 }, { "epoch": 2.2535211267605635, "grad_norm": 4.580643653869629, "learning_rate": 9.859154929577466e-06, "loss": 0.0911, "step": 1760 }, { "epoch": 2.266325224071703, "grad_norm": 10.955920219421387, "learning_rate": 9.852041542182388e-06, "loss": 0.0874, "step": 1770 }, { "epoch": 2.2791293213828423, "grad_norm": 0.017683828249573708, "learning_rate": 9.84492815478731e-06, "loss": 0.0606, "step": 1780 }, { "epoch": 2.291933418693982, "grad_norm": 16.020793914794922, "learning_rate": 9.837814767392234e-06, "loss": 0.0716, "step": 1790 }, { "epoch": 2.3047375160051216, "grad_norm": 8.751802444458008, "learning_rate": 9.830701379997156e-06, "loss": 0.169, "step": 1800 }, { "epoch": 2.317541613316261, "grad_norm": 5.427897930145264, "learning_rate": 9.823587992602078e-06, "loss": 0.119, "step": 1810 }, { "epoch": 2.330345710627401, "grad_norm": 21.988292694091797, "learning_rate": 9.816474605207e-06, "loss": 0.1021, "step": 1820 }, { "epoch": 2.3431498079385404, "grad_norm": 14.013890266418457, "learning_rate": 9.809361217811922e-06, "loss": 0.0747, "step": 1830 }, { "epoch": 2.35595390524968, "grad_norm": 2.2078402042388916, "learning_rate": 9.802247830416845e-06, "loss": 0.2043, "step": 1840 }, { "epoch": 2.3687580025608197, "grad_norm": 0.1353970766067505, "learning_rate": 9.795134443021767e-06, "loss": 0.1157, "step": 1850 }, { "epoch": 2.381562099871959, "grad_norm": 12.870859146118164, "learning_rate": 9.788021055626689e-06, "loss": 0.1143, "step": 1860 }, { "epoch": 2.3943661971830985, "grad_norm": 2.4426429271698, "learning_rate": 9.780907668231613e-06, "loss": 0.1154, "step": 1870 }, { "epoch": 2.407170294494238, "grad_norm": 9.61861515045166, "learning_rate": 9.773794280836535e-06, "loss": 0.0893, "step": 1880 }, { "epoch": 2.4199743918053778, "grad_norm": 5.348551273345947, "learning_rate": 9.766680893441458e-06, "loss": 0.09, "step": 1890 }, { "epoch": 2.4327784891165174, "grad_norm": 11.90121078491211, "learning_rate": 9.75956750604638e-06, "loss": 0.0453, "step": 1900 }, { "epoch": 2.445582586427657, "grad_norm": 7.127316474914551, "learning_rate": 9.752454118651302e-06, "loss": 0.1535, "step": 1910 }, { "epoch": 2.458386683738796, "grad_norm": 13.341761589050293, "learning_rate": 9.745340731256226e-06, "loss": 0.0935, "step": 1920 }, { "epoch": 2.471190781049936, "grad_norm": 7.435236930847168, "learning_rate": 9.738227343861148e-06, "loss": 0.1554, "step": 1930 }, { "epoch": 2.4839948783610755, "grad_norm": 13.538612365722656, "learning_rate": 9.73111395646607e-06, "loss": 0.1084, "step": 1940 }, { "epoch": 2.496798975672215, "grad_norm": 6.2648396492004395, "learning_rate": 9.724000569070993e-06, "loss": 0.1028, "step": 1950 }, { "epoch": 2.5096030729833547, "grad_norm": 10.203313827514648, "learning_rate": 9.716887181675915e-06, "loss": 0.131, "step": 1960 }, { "epoch": 2.5224071702944943, "grad_norm": 9.135252952575684, "learning_rate": 9.709773794280837e-06, "loss": 0.1104, "step": 1970 }, { "epoch": 2.535211267605634, "grad_norm": 1.1023449897766113, "learning_rate": 9.70266040688576e-06, "loss": 0.1732, "step": 1980 }, { "epoch": 2.5480153649167736, "grad_norm": 13.248714447021484, "learning_rate": 9.695547019490681e-06, "loss": 0.0908, "step": 1990 }, { "epoch": 2.560819462227913, "grad_norm": 0.6284473538398743, "learning_rate": 9.688433632095605e-06, "loss": 0.0768, "step": 2000 }, { "epoch": 2.5736235595390524, "grad_norm": 13.200129508972168, "learning_rate": 9.681320244700527e-06, "loss": 0.0869, "step": 2010 }, { "epoch": 2.586427656850192, "grad_norm": 20.896455764770508, "learning_rate": 9.674206857305449e-06, "loss": 0.1065, "step": 2020 }, { "epoch": 2.5992317541613317, "grad_norm": 8.198634147644043, "learning_rate": 9.667093469910372e-06, "loss": 0.1204, "step": 2030 }, { "epoch": 2.6120358514724713, "grad_norm": 3.88381290435791, "learning_rate": 9.659980082515294e-06, "loss": 0.172, "step": 2040 }, { "epoch": 2.6248399487836105, "grad_norm": 8.779280662536621, "learning_rate": 9.652866695120216e-06, "loss": 0.0744, "step": 2050 }, { "epoch": 2.63764404609475, "grad_norm": 7.022649765014648, "learning_rate": 9.64575330772514e-06, "loss": 0.0948, "step": 2060 }, { "epoch": 2.6504481434058897, "grad_norm": 7.353697776794434, "learning_rate": 9.638639920330062e-06, "loss": 0.0541, "step": 2070 }, { "epoch": 2.6632522407170294, "grad_norm": 26.910659790039062, "learning_rate": 9.631526532934986e-06, "loss": 0.1464, "step": 2080 }, { "epoch": 2.676056338028169, "grad_norm": 16.983675003051758, "learning_rate": 9.624413145539908e-06, "loss": 0.1549, "step": 2090 }, { "epoch": 2.6888604353393086, "grad_norm": 9.163124084472656, "learning_rate": 9.61729975814483e-06, "loss": 0.0484, "step": 2100 }, { "epoch": 2.7016645326504483, "grad_norm": 8.896761894226074, "learning_rate": 9.610186370749751e-06, "loss": 0.105, "step": 2110 }, { "epoch": 2.714468629961588, "grad_norm": 7.696724891662598, "learning_rate": 9.603072983354675e-06, "loss": 0.063, "step": 2120 }, { "epoch": 2.7272727272727275, "grad_norm": 10.217194557189941, "learning_rate": 9.595959595959597e-06, "loss": 0.0547, "step": 2130 }, { "epoch": 2.7400768245838667, "grad_norm": 7.503939151763916, "learning_rate": 9.588846208564519e-06, "loss": 0.0765, "step": 2140 }, { "epoch": 2.7528809218950063, "grad_norm": 1.991357684135437, "learning_rate": 9.581732821169441e-06, "loss": 0.0788, "step": 2150 }, { "epoch": 2.765685019206146, "grad_norm": 1.005223035812378, "learning_rate": 9.574619433774363e-06, "loss": 0.1225, "step": 2160 }, { "epoch": 2.7784891165172856, "grad_norm": 11.427959442138672, "learning_rate": 9.567506046379287e-06, "loss": 0.0396, "step": 2170 }, { "epoch": 2.791293213828425, "grad_norm": 5.700253009796143, "learning_rate": 9.560392658984208e-06, "loss": 0.1151, "step": 2180 }, { "epoch": 2.8040973111395644, "grad_norm": 0.054455310106277466, "learning_rate": 9.553279271589132e-06, "loss": 0.1053, "step": 2190 }, { "epoch": 2.816901408450704, "grad_norm": 10.435604095458984, "learning_rate": 9.546165884194054e-06, "loss": 0.102, "step": 2200 }, { "epoch": 2.8297055057618437, "grad_norm": 0.0015482222661376, "learning_rate": 9.539052496798976e-06, "loss": 0.0741, "step": 2210 }, { "epoch": 2.8425096030729833, "grad_norm": 11.852350234985352, "learning_rate": 9.5319391094039e-06, "loss": 0.1502, "step": 2220 }, { "epoch": 2.855313700384123, "grad_norm": 9.063465118408203, "learning_rate": 9.524825722008822e-06, "loss": 0.1396, "step": 2230 }, { "epoch": 2.8681177976952625, "grad_norm": 9.847432136535645, "learning_rate": 9.517712334613744e-06, "loss": 0.0908, "step": 2240 }, { "epoch": 2.880921895006402, "grad_norm": 4.279153347015381, "learning_rate": 9.510598947218667e-06, "loss": 0.074, "step": 2250 }, { "epoch": 2.893725992317542, "grad_norm": 0.3927134871482849, "learning_rate": 9.50348555982359e-06, "loss": 0.1105, "step": 2260 }, { "epoch": 2.9065300896286814, "grad_norm": 6.676069736480713, "learning_rate": 9.496372172428511e-06, "loss": 0.06, "step": 2270 }, { "epoch": 2.9193341869398206, "grad_norm": 20.303152084350586, "learning_rate": 9.489258785033433e-06, "loss": 0.1677, "step": 2280 }, { "epoch": 2.9321382842509602, "grad_norm": 6.692015171051025, "learning_rate": 9.482145397638357e-06, "loss": 0.0588, "step": 2290 }, { "epoch": 2.9449423815621, "grad_norm": 5.088359355926514, "learning_rate": 9.475032010243279e-06, "loss": 0.0945, "step": 2300 }, { "epoch": 2.9577464788732395, "grad_norm": 1.3471933603286743, "learning_rate": 9.4679186228482e-06, "loss": 0.064, "step": 2310 }, { "epoch": 2.970550576184379, "grad_norm": 1.208173155784607, "learning_rate": 9.460805235453123e-06, "loss": 0.0468, "step": 2320 }, { "epoch": 2.9833546734955183, "grad_norm": 10.555696487426758, "learning_rate": 9.453691848058046e-06, "loss": 0.0935, "step": 2330 }, { "epoch": 2.996158770806658, "grad_norm": 0.0008771463762968779, "learning_rate": 9.446578460662968e-06, "loss": 0.1157, "step": 2340 }, { "epoch": 3.0089628681177976, "grad_norm": 10.6071138381958, "learning_rate": 9.43946507326789e-06, "loss": 0.0872, "step": 2350 }, { "epoch": 3.021766965428937, "grad_norm": 7.2021684646606445, "learning_rate": 9.432351685872814e-06, "loss": 0.0657, "step": 2360 }, { "epoch": 3.034571062740077, "grad_norm": 13.131150245666504, "learning_rate": 9.425238298477736e-06, "loss": 0.0751, "step": 2370 }, { "epoch": 3.0473751600512164, "grad_norm": 5.682553768157959, "learning_rate": 9.418124911082658e-06, "loss": 0.1373, "step": 2380 }, { "epoch": 3.060179257362356, "grad_norm": 5.313316345214844, "learning_rate": 9.411011523687581e-06, "loss": 0.037, "step": 2390 }, { "epoch": 3.0729833546734957, "grad_norm": 3.7893171310424805, "learning_rate": 9.403898136292503e-06, "loss": 0.061, "step": 2400 }, { "epoch": 3.085787451984635, "grad_norm": 10.84199333190918, "learning_rate": 9.396784748897427e-06, "loss": 0.0974, "step": 2410 }, { "epoch": 3.0985915492957745, "grad_norm": 4.416197299957275, "learning_rate": 9.389671361502349e-06, "loss": 0.0984, "step": 2420 }, { "epoch": 3.111395646606914, "grad_norm": 0.1039661169052124, "learning_rate": 9.38255797410727e-06, "loss": 0.0483, "step": 2430 }, { "epoch": 3.1241997439180538, "grad_norm": 1.9103771448135376, "learning_rate": 9.375444586712193e-06, "loss": 0.0601, "step": 2440 }, { "epoch": 3.1370038412291934, "grad_norm": 6.648944854736328, "learning_rate": 9.368331199317116e-06, "loss": 0.069, "step": 2450 }, { "epoch": 3.149807938540333, "grad_norm": 7.846061635063961e-05, "learning_rate": 9.361217811922038e-06, "loss": 0.0653, "step": 2460 }, { "epoch": 3.1626120358514727, "grad_norm": 0.6106330156326294, "learning_rate": 9.35410442452696e-06, "loss": 0.0639, "step": 2470 }, { "epoch": 3.175416133162612, "grad_norm": 8.982917785644531, "learning_rate": 9.346991037131882e-06, "loss": 0.0567, "step": 2480 }, { "epoch": 3.1882202304737515, "grad_norm": 2.2257747650146484, "learning_rate": 9.339877649736804e-06, "loss": 0.076, "step": 2490 }, { "epoch": 3.201024327784891, "grad_norm": 7.568869590759277, "learning_rate": 9.332764262341728e-06, "loss": 0.0271, "step": 2500 }, { "epoch": 3.2138284250960307, "grad_norm": 11.3545560836792, "learning_rate": 9.32565087494665e-06, "loss": 0.0405, "step": 2510 }, { "epoch": 3.2266325224071704, "grad_norm": 0.000226711985305883, "learning_rate": 9.318537487551573e-06, "loss": 0.0552, "step": 2520 }, { "epoch": 3.23943661971831, "grad_norm": 7.903006553649902, "learning_rate": 9.311424100156495e-06, "loss": 0.0689, "step": 2530 }, { "epoch": 3.2522407170294496, "grad_norm": 5.433798789978027, "learning_rate": 9.304310712761417e-06, "loss": 0.0527, "step": 2540 }, { "epoch": 3.265044814340589, "grad_norm": 9.897836685180664, "learning_rate": 9.297197325366341e-06, "loss": 0.1249, "step": 2550 }, { "epoch": 3.2778489116517284, "grad_norm": 5.51058292388916, "learning_rate": 9.290083937971263e-06, "loss": 0.0486, "step": 2560 }, { "epoch": 3.290653008962868, "grad_norm": 14.336811065673828, "learning_rate": 9.282970550576185e-06, "loss": 0.0988, "step": 2570 }, { "epoch": 3.3034571062740077, "grad_norm": 7.548839092254639, "learning_rate": 9.275857163181108e-06, "loss": 0.0882, "step": 2580 }, { "epoch": 3.3162612035851473, "grad_norm": 2.035888910293579, "learning_rate": 9.26874377578603e-06, "loss": 0.0445, "step": 2590 }, { "epoch": 3.329065300896287, "grad_norm": 12.386154174804688, "learning_rate": 9.261630388390952e-06, "loss": 0.12, "step": 2600 }, { "epoch": 3.3418693982074266, "grad_norm": 6.706932067871094, "learning_rate": 9.254517000995874e-06, "loss": 0.0787, "step": 2610 }, { "epoch": 3.3546734955185658, "grad_norm": 0.008335432037711143, "learning_rate": 9.247403613600798e-06, "loss": 0.075, "step": 2620 }, { "epoch": 3.3674775928297054, "grad_norm": 0.6694245934486389, "learning_rate": 9.24029022620572e-06, "loss": 0.0871, "step": 2630 }, { "epoch": 3.380281690140845, "grad_norm": 0.02344265952706337, "learning_rate": 9.233176838810642e-06, "loss": 0.0611, "step": 2640 }, { "epoch": 3.3930857874519846, "grad_norm": 0.15562568604946136, "learning_rate": 9.226063451415564e-06, "loss": 0.0467, "step": 2650 }, { "epoch": 3.4058898847631243, "grad_norm": 11.615291595458984, "learning_rate": 9.218950064020487e-06, "loss": 0.0701, "step": 2660 }, { "epoch": 3.418693982074264, "grad_norm": 1.564645767211914, "learning_rate": 9.21183667662541e-06, "loss": 0.0562, "step": 2670 }, { "epoch": 3.4314980793854035, "grad_norm": 0.8504931330680847, "learning_rate": 9.204723289230331e-06, "loss": 0.0563, "step": 2680 }, { "epoch": 3.4443021766965427, "grad_norm": 8.873530387878418, "learning_rate": 9.197609901835255e-06, "loss": 0.0563, "step": 2690 }, { "epoch": 3.4571062740076823, "grad_norm": 7.516530513763428, "learning_rate": 9.190496514440177e-06, "loss": 0.0638, "step": 2700 }, { "epoch": 3.469910371318822, "grad_norm": 10.838394165039062, "learning_rate": 9.1833831270451e-06, "loss": 0.0916, "step": 2710 }, { "epoch": 3.4827144686299616, "grad_norm": 12.697843551635742, "learning_rate": 9.176269739650023e-06, "loss": 0.0794, "step": 2720 }, { "epoch": 3.495518565941101, "grad_norm": 4.907537937164307, "learning_rate": 9.169156352254945e-06, "loss": 0.0672, "step": 2730 }, { "epoch": 3.508322663252241, "grad_norm": 8.670528411865234, "learning_rate": 9.162042964859868e-06, "loss": 0.101, "step": 2740 }, { "epoch": 3.52112676056338, "grad_norm": 0.005142922047525644, "learning_rate": 9.15492957746479e-06, "loss": 0.0832, "step": 2750 }, { "epoch": 3.5339308578745197, "grad_norm": 7.155096054077148, "learning_rate": 9.147816190069712e-06, "loss": 0.0617, "step": 2760 }, { "epoch": 3.5467349551856593, "grad_norm": 8.930415153503418, "learning_rate": 9.140702802674634e-06, "loss": 0.1017, "step": 2770 }, { "epoch": 3.559539052496799, "grad_norm": 12.03281021118164, "learning_rate": 9.133589415279556e-06, "loss": 0.0663, "step": 2780 }, { "epoch": 3.5723431498079385, "grad_norm": 4.852962970733643, "learning_rate": 9.12647602788448e-06, "loss": 0.0531, "step": 2790 }, { "epoch": 3.585147247119078, "grad_norm": 4.8767571449279785, "learning_rate": 9.119362640489402e-06, "loss": 0.0693, "step": 2800 }, { "epoch": 3.597951344430218, "grad_norm": 8.036718368530273, "learning_rate": 9.112249253094323e-06, "loss": 0.0757, "step": 2810 }, { "epoch": 3.6107554417413574, "grad_norm": 0.04634200781583786, "learning_rate": 9.105135865699247e-06, "loss": 0.0207, "step": 2820 }, { "epoch": 3.623559539052497, "grad_norm": 2.246762752532959, "learning_rate": 9.098022478304169e-06, "loss": 0.0833, "step": 2830 }, { "epoch": 3.6363636363636362, "grad_norm": 0.8394132256507874, "learning_rate": 9.090909090909091e-06, "loss": 0.0546, "step": 2840 }, { "epoch": 3.649167733674776, "grad_norm": 0.859449565410614, "learning_rate": 9.083795703514015e-06, "loss": 0.0529, "step": 2850 }, { "epoch": 3.6619718309859155, "grad_norm": 8.246654510498047, "learning_rate": 9.076682316118937e-06, "loss": 0.0732, "step": 2860 }, { "epoch": 3.674775928297055, "grad_norm": 8.581944465637207, "learning_rate": 9.069568928723859e-06, "loss": 0.0841, "step": 2870 }, { "epoch": 3.6875800256081948, "grad_norm": 6.378680229187012, "learning_rate": 9.062455541328782e-06, "loss": 0.054, "step": 2880 }, { "epoch": 3.700384122919334, "grad_norm": 0.03322785347700119, "learning_rate": 9.055342153933704e-06, "loss": 0.0722, "step": 2890 }, { "epoch": 3.7131882202304736, "grad_norm": 6.124284744262695, "learning_rate": 9.048228766538626e-06, "loss": 0.0479, "step": 2900 }, { "epoch": 3.725992317541613, "grad_norm": 0.22712665796279907, "learning_rate": 9.04111537914355e-06, "loss": 0.0476, "step": 2910 }, { "epoch": 3.738796414852753, "grad_norm": 6.798972129821777, "learning_rate": 9.034001991748472e-06, "loss": 0.073, "step": 2920 }, { "epoch": 3.7516005121638925, "grad_norm": 0.0032643969170749187, "learning_rate": 9.026888604353394e-06, "loss": 0.0111, "step": 2930 }, { "epoch": 3.764404609475032, "grad_norm": 11.23230266571045, "learning_rate": 9.019775216958316e-06, "loss": 0.0706, "step": 2940 }, { "epoch": 3.7772087067861717, "grad_norm": 4.730894565582275, "learning_rate": 9.01266182956324e-06, "loss": 0.0324, "step": 2950 }, { "epoch": 3.7900128040973113, "grad_norm": 7.80692195892334, "learning_rate": 9.005548442168161e-06, "loss": 0.1034, "step": 2960 }, { "epoch": 3.802816901408451, "grad_norm": 18.416215896606445, "learning_rate": 8.998435054773083e-06, "loss": 0.1731, "step": 2970 }, { "epoch": 3.81562099871959, "grad_norm": 6.270477771759033, "learning_rate": 8.991321667378005e-06, "loss": 0.0642, "step": 2980 }, { "epoch": 3.82842509603073, "grad_norm": 2.932262659072876, "learning_rate": 8.984208279982929e-06, "loss": 0.0817, "step": 2990 }, { "epoch": 3.8412291933418694, "grad_norm": 8.80971908569336, "learning_rate": 8.97709489258785e-06, "loss": 0.0743, "step": 3000 }, { "epoch": 3.854033290653009, "grad_norm": 11.367161750793457, "learning_rate": 8.969981505192773e-06, "loss": 0.0502, "step": 3010 }, { "epoch": 3.8668373879641487, "grad_norm": 0.002171382075175643, "learning_rate": 8.962868117797696e-06, "loss": 0.0641, "step": 3020 }, { "epoch": 3.879641485275288, "grad_norm": 0.4532616436481476, "learning_rate": 8.955754730402618e-06, "loss": 0.0651, "step": 3030 }, { "epoch": 3.8924455825864275, "grad_norm": 0.7072662711143494, "learning_rate": 8.948641343007542e-06, "loss": 0.0793, "step": 3040 }, { "epoch": 3.905249679897567, "grad_norm": 6.481967449188232, "learning_rate": 8.941527955612464e-06, "loss": 0.0815, "step": 3050 }, { "epoch": 3.9180537772087067, "grad_norm": 0.0189608633518219, "learning_rate": 8.934414568217386e-06, "loss": 0.0957, "step": 3060 }, { "epoch": 3.9308578745198464, "grad_norm": 9.144214630126953, "learning_rate": 8.92730118082231e-06, "loss": 0.0575, "step": 3070 }, { "epoch": 3.943661971830986, "grad_norm": 9.871874809265137, "learning_rate": 8.920187793427231e-06, "loss": 0.1127, "step": 3080 }, { "epoch": 3.9564660691421256, "grad_norm": 9.677109718322754, "learning_rate": 8.913074406032153e-06, "loss": 0.0978, "step": 3090 }, { "epoch": 3.9692701664532652, "grad_norm": 3.6317031383514404, "learning_rate": 8.905961018637075e-06, "loss": 0.0738, "step": 3100 }, { "epoch": 3.982074263764405, "grad_norm": 1.407476544380188, "learning_rate": 8.898847631241997e-06, "loss": 0.0363, "step": 3110 }, { "epoch": 3.994878361075544, "grad_norm": 7.945038318634033, "learning_rate": 8.891734243846921e-06, "loss": 0.074, "step": 3120 }, { "epoch": 4.007682458386684, "grad_norm": 0.7004364132881165, "learning_rate": 8.884620856451843e-06, "loss": 0.0445, "step": 3130 }, { "epoch": 4.020486555697823, "grad_norm": 8.339925765991211, "learning_rate": 8.877507469056765e-06, "loss": 0.0311, "step": 3140 }, { "epoch": 4.0332906530089625, "grad_norm": 8.701236724853516, "learning_rate": 8.870394081661688e-06, "loss": 0.0563, "step": 3150 }, { "epoch": 4.046094750320102, "grad_norm": 11.493291854858398, "learning_rate": 8.86328069426661e-06, "loss": 0.1014, "step": 3160 }, { "epoch": 4.058898847631242, "grad_norm": 14.968050956726074, "learning_rate": 8.856167306871532e-06, "loss": 0.0572, "step": 3170 }, { "epoch": 4.071702944942381, "grad_norm": 6.135365009307861, "learning_rate": 8.849053919476456e-06, "loss": 0.0502, "step": 3180 }, { "epoch": 4.084507042253521, "grad_norm": 3.1343040466308594, "learning_rate": 8.841940532081378e-06, "loss": 0.0305, "step": 3190 }, { "epoch": 4.097311139564661, "grad_norm": 7.629443168640137, "learning_rate": 8.8348271446863e-06, "loss": 0.0682, "step": 3200 }, { "epoch": 4.1101152368758, "grad_norm": 7.97960090637207, "learning_rate": 8.827713757291224e-06, "loss": 0.0786, "step": 3210 }, { "epoch": 4.12291933418694, "grad_norm": 1.2798930406570435, "learning_rate": 8.820600369896145e-06, "loss": 0.043, "step": 3220 }, { "epoch": 4.1357234314980795, "grad_norm": 0.1385413408279419, "learning_rate": 8.813486982501069e-06, "loss": 0.049, "step": 3230 }, { "epoch": 4.148527528809219, "grad_norm": 1.2011138200759888, "learning_rate": 8.806373595105991e-06, "loss": 0.0859, "step": 3240 }, { "epoch": 4.161331626120359, "grad_norm": 7.076838970184326, "learning_rate": 8.799260207710913e-06, "loss": 0.0635, "step": 3250 }, { "epoch": 4.174135723431498, "grad_norm": 0.1687900722026825, "learning_rate": 8.792146820315835e-06, "loss": 0.0686, "step": 3260 }, { "epoch": 4.186939820742638, "grad_norm": 3.7487645149230957, "learning_rate": 8.785033432920757e-06, "loss": 0.0468, "step": 3270 }, { "epoch": 4.199743918053777, "grad_norm": 0.12559740245342255, "learning_rate": 8.777920045525679e-06, "loss": 0.0403, "step": 3280 }, { "epoch": 4.212548015364916, "grad_norm": 0.34440168738365173, "learning_rate": 8.770806658130602e-06, "loss": 0.0867, "step": 3290 }, { "epoch": 4.225352112676056, "grad_norm": 0.0011506618466228247, "learning_rate": 8.763693270735524e-06, "loss": 0.0354, "step": 3300 }, { "epoch": 4.238156209987196, "grad_norm": 9.541993141174316, "learning_rate": 8.756579883340446e-06, "loss": 0.0473, "step": 3310 }, { "epoch": 4.250960307298335, "grad_norm": 0.023387501016259193, "learning_rate": 8.74946649594537e-06, "loss": 0.0605, "step": 3320 }, { "epoch": 4.263764404609475, "grad_norm": 7.409848690032959, "learning_rate": 8.742353108550292e-06, "loss": 0.0402, "step": 3330 }, { "epoch": 4.2765685019206146, "grad_norm": 7.853289604187012, "learning_rate": 8.735239721155216e-06, "loss": 0.0535, "step": 3340 }, { "epoch": 4.289372599231754, "grad_norm": 10.809486389160156, "learning_rate": 8.728126333760138e-06, "loss": 0.0493, "step": 3350 }, { "epoch": 4.302176696542894, "grad_norm": 0.22432303428649902, "learning_rate": 8.72101294636506e-06, "loss": 0.0255, "step": 3360 }, { "epoch": 4.314980793854033, "grad_norm": 7.155245780944824, "learning_rate": 8.713899558969983e-06, "loss": 0.0362, "step": 3370 }, { "epoch": 4.327784891165173, "grad_norm": 0.5234391689300537, "learning_rate": 8.706786171574905e-06, "loss": 0.0559, "step": 3380 }, { "epoch": 4.340588988476313, "grad_norm": 6.478000164031982, "learning_rate": 8.699672784179827e-06, "loss": 0.0475, "step": 3390 }, { "epoch": 4.353393085787452, "grad_norm": 7.4332709312438965, "learning_rate": 8.69255939678475e-06, "loss": 0.028, "step": 3400 }, { "epoch": 4.366197183098592, "grad_norm": 0.00016584369586780667, "learning_rate": 8.685446009389673e-06, "loss": 0.033, "step": 3410 }, { "epoch": 4.379001280409732, "grad_norm": 5.26579475402832, "learning_rate": 8.678332621994595e-06, "loss": 0.0266, "step": 3420 }, { "epoch": 4.39180537772087, "grad_norm": 0.01609189808368683, "learning_rate": 8.671219234599517e-06, "loss": 0.052, "step": 3430 }, { "epoch": 4.40460947503201, "grad_norm": 0.04848508909344673, "learning_rate": 8.664105847204439e-06, "loss": 0.0623, "step": 3440 }, { "epoch": 4.41741357234315, "grad_norm": 0.03006572835147381, "learning_rate": 8.656992459809362e-06, "loss": 0.0358, "step": 3450 }, { "epoch": 4.430217669654289, "grad_norm": 0.03689035773277283, "learning_rate": 8.649879072414284e-06, "loss": 0.0281, "step": 3460 }, { "epoch": 4.443021766965429, "grad_norm": 5.47698450088501, "learning_rate": 8.642765685019206e-06, "loss": 0.0685, "step": 3470 }, { "epoch": 4.4558258642765685, "grad_norm": 7.300468921661377, "learning_rate": 8.63565229762413e-06, "loss": 0.0607, "step": 3480 }, { "epoch": 4.468629961587708, "grad_norm": 5.875164985656738, "learning_rate": 8.628538910229052e-06, "loss": 0.0694, "step": 3490 }, { "epoch": 4.481434058898848, "grad_norm": 11.426108360290527, "learning_rate": 8.621425522833974e-06, "loss": 0.0783, "step": 3500 }, { "epoch": 4.494238156209987, "grad_norm": 9.112909317016602, "learning_rate": 8.614312135438897e-06, "loss": 0.0664, "step": 3510 }, { "epoch": 4.507042253521127, "grad_norm": 0.24784012138843536, "learning_rate": 8.60719874804382e-06, "loss": 0.0124, "step": 3520 }, { "epoch": 4.519846350832267, "grad_norm": 9.857042459771037e-05, "learning_rate": 8.600085360648741e-06, "loss": 0.0428, "step": 3530 }, { "epoch": 4.532650448143406, "grad_norm": 8.992209434509277, "learning_rate": 8.592971973253665e-06, "loss": 0.0435, "step": 3540 }, { "epoch": 4.545454545454545, "grad_norm": 6.475641250610352, "learning_rate": 8.585858585858587e-06, "loss": 0.0547, "step": 3550 }, { "epoch": 4.558258642765685, "grad_norm": 14.07235336303711, "learning_rate": 8.578745198463509e-06, "loss": 0.0467, "step": 3560 }, { "epoch": 4.571062740076824, "grad_norm": 7.2002854347229, "learning_rate": 8.571631811068432e-06, "loss": 0.0091, "step": 3570 }, { "epoch": 4.583866837387964, "grad_norm": 0.07768731564283371, "learning_rate": 8.564518423673354e-06, "loss": 0.0467, "step": 3580 }, { "epoch": 4.5966709346991035, "grad_norm": 11.106244087219238, "learning_rate": 8.557405036278276e-06, "loss": 0.0477, "step": 3590 }, { "epoch": 4.609475032010243, "grad_norm": 0.0894252136349678, "learning_rate": 8.550291648883198e-06, "loss": 0.0285, "step": 3600 }, { "epoch": 4.622279129321383, "grad_norm": 5.591623783111572, "learning_rate": 8.54317826148812e-06, "loss": 0.0706, "step": 3610 }, { "epoch": 4.635083226632522, "grad_norm": 0.042741406708955765, "learning_rate": 8.536064874093044e-06, "loss": 0.0221, "step": 3620 }, { "epoch": 4.647887323943662, "grad_norm": 3.735109329223633, "learning_rate": 8.528951486697966e-06, "loss": 0.043, "step": 3630 }, { "epoch": 4.660691421254802, "grad_norm": 4.355711936950684, "learning_rate": 8.521838099302888e-06, "loss": 0.0725, "step": 3640 }, { "epoch": 4.673495518565941, "grad_norm": 6.873217582702637, "learning_rate": 8.514724711907811e-06, "loss": 0.0604, "step": 3650 }, { "epoch": 4.686299615877081, "grad_norm": 9.788108825683594, "learning_rate": 8.507611324512733e-06, "loss": 0.0407, "step": 3660 }, { "epoch": 4.6991037131882205, "grad_norm": 9.999732971191406, "learning_rate": 8.500497937117657e-06, "loss": 0.0821, "step": 3670 }, { "epoch": 4.71190781049936, "grad_norm": 0.6053422093391418, "learning_rate": 8.493384549722579e-06, "loss": 0.0256, "step": 3680 }, { "epoch": 4.7247119078105, "grad_norm": 9.958725929260254, "learning_rate": 8.4862711623275e-06, "loss": 0.0413, "step": 3690 }, { "epoch": 4.737516005121639, "grad_norm": 7.323482513427734, "learning_rate": 8.479157774932424e-06, "loss": 0.0379, "step": 3700 }, { "epoch": 4.750320102432778, "grad_norm": 3.1788806915283203, "learning_rate": 8.472044387537346e-06, "loss": 0.0652, "step": 3710 }, { "epoch": 4.763124199743918, "grad_norm": 7.410254955291748, "learning_rate": 8.464931000142268e-06, "loss": 0.1127, "step": 3720 }, { "epoch": 4.775928297055057, "grad_norm": 15.534052848815918, "learning_rate": 8.457817612747192e-06, "loss": 0.0765, "step": 3730 }, { "epoch": 4.788732394366197, "grad_norm": 2.075538158416748, "learning_rate": 8.450704225352114e-06, "loss": 0.0589, "step": 3740 }, { "epoch": 4.801536491677337, "grad_norm": 11.67330551147461, "learning_rate": 8.443590837957036e-06, "loss": 0.0716, "step": 3750 }, { "epoch": 4.814340588988476, "grad_norm": 6.3938798904418945, "learning_rate": 8.436477450561958e-06, "loss": 0.0182, "step": 3760 }, { "epoch": 4.827144686299616, "grad_norm": 6.382133960723877, "learning_rate": 8.42936406316688e-06, "loss": 0.0634, "step": 3770 }, { "epoch": 4.8399487836107555, "grad_norm": 6.629985809326172, "learning_rate": 8.422250675771803e-06, "loss": 0.0913, "step": 3780 }, { "epoch": 4.852752880921895, "grad_norm": 0.031200766563415527, "learning_rate": 8.415137288376725e-06, "loss": 0.0359, "step": 3790 }, { "epoch": 4.865556978233035, "grad_norm": 2.89249587059021, "learning_rate": 8.408023900981647e-06, "loss": 0.0178, "step": 3800 }, { "epoch": 4.878361075544174, "grad_norm": 7.646574974060059, "learning_rate": 8.400910513586571e-06, "loss": 0.0314, "step": 3810 }, { "epoch": 4.891165172855314, "grad_norm": 6.414608001708984, "learning_rate": 8.393797126191493e-06, "loss": 0.0833, "step": 3820 }, { "epoch": 4.903969270166453, "grad_norm": 1.906454086303711, "learning_rate": 8.386683738796415e-06, "loss": 0.0654, "step": 3830 }, { "epoch": 4.916773367477592, "grad_norm": 12.821189880371094, "learning_rate": 8.379570351401339e-06, "loss": 0.0852, "step": 3840 }, { "epoch": 4.929577464788732, "grad_norm": 5.904670715332031, "learning_rate": 8.37245696400626e-06, "loss": 0.0496, "step": 3850 }, { "epoch": 4.942381562099872, "grad_norm": 13.252474784851074, "learning_rate": 8.365343576611184e-06, "loss": 0.0373, "step": 3860 }, { "epoch": 4.955185659411011, "grad_norm": 9.706033706665039, "learning_rate": 8.358230189216106e-06, "loss": 0.1003, "step": 3870 }, { "epoch": 4.967989756722151, "grad_norm": 5.731202602386475, "learning_rate": 8.351116801821028e-06, "loss": 0.0162, "step": 3880 }, { "epoch": 4.980793854033291, "grad_norm": 0.10428772121667862, "learning_rate": 8.34400341442595e-06, "loss": 0.0549, "step": 3890 }, { "epoch": 4.99359795134443, "grad_norm": 3.261608123779297, "learning_rate": 8.336890027030874e-06, "loss": 0.0289, "step": 3900 }, { "epoch": 5.00640204865557, "grad_norm": 0.6107264757156372, "learning_rate": 8.329776639635796e-06, "loss": 0.0824, "step": 3910 }, { "epoch": 5.019206145966709, "grad_norm": 4.616156578063965, "learning_rate": 8.322663252240718e-06, "loss": 0.0455, "step": 3920 }, { "epoch": 5.032010243277849, "grad_norm": 5.969337463378906, "learning_rate": 8.31554986484564e-06, "loss": 0.0449, "step": 3930 }, { "epoch": 5.044814340588989, "grad_norm": 0.1294054538011551, "learning_rate": 8.308436477450561e-06, "loss": 0.0158, "step": 3940 }, { "epoch": 5.057618437900128, "grad_norm": 7.5108184814453125, "learning_rate": 8.301323090055485e-06, "loss": 0.0352, "step": 3950 }, { "epoch": 5.070422535211268, "grad_norm": 8.7492036819458, "learning_rate": 8.294209702660407e-06, "loss": 0.031, "step": 3960 }, { "epoch": 5.083226632522408, "grad_norm": 2.8963265419006348, "learning_rate": 8.28709631526533e-06, "loss": 0.0178, "step": 3970 }, { "epoch": 5.096030729833546, "grad_norm": 7.99992561340332, "learning_rate": 8.279982927870253e-06, "loss": 0.0145, "step": 3980 }, { "epoch": 5.108834827144686, "grad_norm": 0.007856784388422966, "learning_rate": 8.272869540475175e-06, "loss": 0.0431, "step": 3990 }, { "epoch": 5.121638924455826, "grad_norm": 7.905889511108398, "learning_rate": 8.265756153080098e-06, "loss": 0.0547, "step": 4000 }, { "epoch": 5.134443021766965, "grad_norm": 6.235837936401367, "learning_rate": 8.25864276568502e-06, "loss": 0.0479, "step": 4010 }, { "epoch": 5.147247119078105, "grad_norm": 0.536173939704895, "learning_rate": 8.251529378289942e-06, "loss": 0.0274, "step": 4020 }, { "epoch": 5.1600512163892445, "grad_norm": 8.95507526397705, "learning_rate": 8.244415990894866e-06, "loss": 0.0629, "step": 4030 }, { "epoch": 5.172855313700384, "grad_norm": 0.33378151059150696, "learning_rate": 8.237302603499788e-06, "loss": 0.0984, "step": 4040 }, { "epoch": 5.185659411011524, "grad_norm": 0.033245958387851715, "learning_rate": 8.23018921610471e-06, "loss": 0.0414, "step": 4050 }, { "epoch": 5.198463508322663, "grad_norm": 0.03647937998175621, "learning_rate": 8.223075828709632e-06, "loss": 0.0165, "step": 4060 }, { "epoch": 5.211267605633803, "grad_norm": 7.125176906585693, "learning_rate": 8.215962441314555e-06, "loss": 0.048, "step": 4070 }, { "epoch": 5.224071702944943, "grad_norm": 1.0387356281280518, "learning_rate": 8.208849053919477e-06, "loss": 0.0525, "step": 4080 }, { "epoch": 5.236875800256082, "grad_norm": 6.314057350158691, "learning_rate": 8.201735666524399e-06, "loss": 0.0265, "step": 4090 }, { "epoch": 5.249679897567222, "grad_norm": 3.168534994125366, "learning_rate": 8.194622279129321e-06, "loss": 0.0402, "step": 4100 }, { "epoch": 5.262483994878361, "grad_norm": 7.815944671630859, "learning_rate": 8.187508891734245e-06, "loss": 0.0653, "step": 4110 }, { "epoch": 5.2752880921895, "grad_norm": 4.802215099334717, "learning_rate": 8.180395504339167e-06, "loss": 0.039, "step": 4120 }, { "epoch": 5.28809218950064, "grad_norm": 0.07285226881504059, "learning_rate": 8.173282116944089e-06, "loss": 0.0305, "step": 4130 }, { "epoch": 5.3008962868117795, "grad_norm": 0.01511402614414692, "learning_rate": 8.166168729549012e-06, "loss": 0.0335, "step": 4140 }, { "epoch": 5.313700384122919, "grad_norm": 0.45437008142471313, "learning_rate": 8.159055342153934e-06, "loss": 0.0783, "step": 4150 }, { "epoch": 5.326504481434059, "grad_norm": 0.014249751344323158, "learning_rate": 8.151941954758856e-06, "loss": 0.0452, "step": 4160 }, { "epoch": 5.339308578745198, "grad_norm": 0.15017516911029816, "learning_rate": 8.14482856736378e-06, "loss": 0.034, "step": 4170 }, { "epoch": 5.352112676056338, "grad_norm": 4.36442232131958, "learning_rate": 8.137715179968702e-06, "loss": 0.0151, "step": 4180 }, { "epoch": 5.364916773367478, "grad_norm": 7.4028544425964355, "learning_rate": 8.130601792573625e-06, "loss": 0.0299, "step": 4190 }, { "epoch": 5.377720870678617, "grad_norm": 3.0723438262939453, "learning_rate": 8.123488405178547e-06, "loss": 0.0361, "step": 4200 }, { "epoch": 5.390524967989757, "grad_norm": 4.94050407409668, "learning_rate": 8.11637501778347e-06, "loss": 0.0413, "step": 4210 }, { "epoch": 5.4033290653008965, "grad_norm": 0.8614751100540161, "learning_rate": 8.109261630388391e-06, "loss": 0.0127, "step": 4220 }, { "epoch": 5.416133162612036, "grad_norm": 10.043490409851074, "learning_rate": 8.102148242993315e-06, "loss": 0.076, "step": 4230 }, { "epoch": 5.428937259923176, "grad_norm": 7.008946895599365, "learning_rate": 8.095034855598237e-06, "loss": 0.0264, "step": 4240 }, { "epoch": 5.441741357234315, "grad_norm": 4.448896884918213, "learning_rate": 8.087921468203159e-06, "loss": 0.0674, "step": 4250 }, { "epoch": 5.454545454545454, "grad_norm": 0.11608394980430603, "learning_rate": 8.08080808080808e-06, "loss": 0.0322, "step": 4260 }, { "epoch": 5.467349551856594, "grad_norm": 4.199873728794046e-05, "learning_rate": 8.073694693413003e-06, "loss": 0.0131, "step": 4270 }, { "epoch": 5.480153649167733, "grad_norm": 8.15583324432373, "learning_rate": 8.066581306017926e-06, "loss": 0.0352, "step": 4280 }, { "epoch": 5.492957746478873, "grad_norm": 8.766875267028809, "learning_rate": 8.059467918622848e-06, "loss": 0.0523, "step": 4290 }, { "epoch": 5.505761843790013, "grad_norm": 4.205753803253174, "learning_rate": 8.052354531227772e-06, "loss": 0.0455, "step": 4300 }, { "epoch": 5.518565941101152, "grad_norm": 9.919289588928223, "learning_rate": 8.045241143832694e-06, "loss": 0.0353, "step": 4310 }, { "epoch": 5.531370038412292, "grad_norm": 0.001328068203292787, "learning_rate": 8.038127756437616e-06, "loss": 0.0346, "step": 4320 }, { "epoch": 5.5441741357234315, "grad_norm": 0.0011309911496937275, "learning_rate": 8.03101436904254e-06, "loss": 0.0601, "step": 4330 }, { "epoch": 5.556978233034571, "grad_norm": 0.8580485582351685, "learning_rate": 8.023900981647461e-06, "loss": 0.039, "step": 4340 }, { "epoch": 5.569782330345711, "grad_norm": 1.1906603574752808, "learning_rate": 8.016787594252383e-06, "loss": 0.0656, "step": 4350 }, { "epoch": 5.58258642765685, "grad_norm": 0.004430301487445831, "learning_rate": 8.009674206857307e-06, "loss": 0.0539, "step": 4360 }, { "epoch": 5.59539052496799, "grad_norm": 9.235508918762207, "learning_rate": 8.002560819462229e-06, "loss": 0.0437, "step": 4370 }, { "epoch": 5.60819462227913, "grad_norm": 7.584328651428223, "learning_rate": 7.995447432067151e-06, "loss": 0.0363, "step": 4380 }, { "epoch": 5.620998719590268, "grad_norm": 0.2141062319278717, "learning_rate": 7.988334044672073e-06, "loss": 0.0396, "step": 4390 }, { "epoch": 5.633802816901408, "grad_norm": 0.028521370142698288, "learning_rate": 7.981220657276996e-06, "loss": 0.0291, "step": 4400 }, { "epoch": 5.646606914212548, "grad_norm": 4.2861199378967285, "learning_rate": 7.974107269881918e-06, "loss": 0.016, "step": 4410 }, { "epoch": 5.659411011523687, "grad_norm": 0.19598886370658875, "learning_rate": 7.96699388248684e-06, "loss": 0.0307, "step": 4420 }, { "epoch": 5.672215108834827, "grad_norm": 0.2010897547006607, "learning_rate": 7.959880495091762e-06, "loss": 0.0112, "step": 4430 }, { "epoch": 5.685019206145967, "grad_norm": 11.772674560546875, "learning_rate": 7.952767107696686e-06, "loss": 0.0638, "step": 4440 }, { "epoch": 5.697823303457106, "grad_norm": 0.010794471949338913, "learning_rate": 7.945653720301608e-06, "loss": 0.0264, "step": 4450 }, { "epoch": 5.710627400768246, "grad_norm": 7.447845935821533, "learning_rate": 7.93854033290653e-06, "loss": 0.0436, "step": 4460 }, { "epoch": 5.7234314980793854, "grad_norm": 0.003293869085609913, "learning_rate": 7.931426945511454e-06, "loss": 0.0326, "step": 4470 }, { "epoch": 5.736235595390525, "grad_norm": 5.008636474609375, "learning_rate": 7.924313558116375e-06, "loss": 0.0408, "step": 4480 }, { "epoch": 5.749039692701665, "grad_norm": 0.0017159223789349198, "learning_rate": 7.917200170721299e-06, "loss": 0.0521, "step": 4490 }, { "epoch": 5.761843790012804, "grad_norm": 0.03760657459497452, "learning_rate": 7.910086783326221e-06, "loss": 0.0237, "step": 4500 }, { "epoch": 5.774647887323944, "grad_norm": 0.9787740111351013, "learning_rate": 7.902973395931143e-06, "loss": 0.0137, "step": 4510 }, { "epoch": 5.787451984635084, "grad_norm": 7.57289457321167, "learning_rate": 7.895860008536067e-06, "loss": 0.0188, "step": 4520 }, { "epoch": 5.800256081946223, "grad_norm": 6.602818012237549, "learning_rate": 7.888746621140989e-06, "loss": 0.0374, "step": 4530 }, { "epoch": 5.813060179257363, "grad_norm": 6.57660436630249, "learning_rate": 7.88163323374591e-06, "loss": 0.033, "step": 4540 }, { "epoch": 5.825864276568502, "grad_norm": 8.515735626220703, "learning_rate": 7.874519846350833e-06, "loss": 0.0393, "step": 4550 }, { "epoch": 5.838668373879641, "grad_norm": 1.3361997604370117, "learning_rate": 7.867406458955754e-06, "loss": 0.0145, "step": 4560 }, { "epoch": 5.851472471190781, "grad_norm": 7.162911415100098, "learning_rate": 7.860293071560678e-06, "loss": 0.0611, "step": 4570 }, { "epoch": 5.8642765685019205, "grad_norm": 11.508686065673828, "learning_rate": 7.8531796841656e-06, "loss": 0.0545, "step": 4580 }, { "epoch": 5.87708066581306, "grad_norm": 9.42324447631836, "learning_rate": 7.846066296770522e-06, "loss": 0.034, "step": 4590 }, { "epoch": 5.8898847631242, "grad_norm": 0.0004297104605939239, "learning_rate": 7.838952909375446e-06, "loss": 0.02, "step": 4600 }, { "epoch": 5.902688860435339, "grad_norm": 0.0026710787788033485, "learning_rate": 7.831839521980368e-06, "loss": 0.0446, "step": 4610 }, { "epoch": 5.915492957746479, "grad_norm": 12.91624927520752, "learning_rate": 7.82472613458529e-06, "loss": 0.0642, "step": 4620 }, { "epoch": 5.928297055057619, "grad_norm": 6.404867172241211, "learning_rate": 7.817612747190213e-06, "loss": 0.0446, "step": 4630 }, { "epoch": 5.941101152368758, "grad_norm": 6.4475860595703125, "learning_rate": 7.810499359795135e-06, "loss": 0.0383, "step": 4640 }, { "epoch": 5.953905249679898, "grad_norm": 10.074686050415039, "learning_rate": 7.803385972400057e-06, "loss": 0.0328, "step": 4650 }, { "epoch": 5.9667093469910375, "grad_norm": 9.27889633178711, "learning_rate": 7.79627258500498e-06, "loss": 0.046, "step": 4660 }, { "epoch": 5.979513444302176, "grad_norm": 8.591060638427734, "learning_rate": 7.789159197609903e-06, "loss": 0.0274, "step": 4670 }, { "epoch": 5.992317541613316, "grad_norm": 0.004389516543596983, "learning_rate": 7.782045810214825e-06, "loss": 0.0398, "step": 4680 }, { "epoch": 6.0051216389244555, "grad_norm": 10.176156044006348, "learning_rate": 7.774932422819748e-06, "loss": 0.0619, "step": 4690 }, { "epoch": 6.017925736235595, "grad_norm": 5.341851711273193, "learning_rate": 7.76781903542467e-06, "loss": 0.017, "step": 4700 }, { "epoch": 6.030729833546735, "grad_norm": 5.233988285064697, "learning_rate": 7.760705648029592e-06, "loss": 0.0358, "step": 4710 }, { "epoch": 6.043533930857874, "grad_norm": 11.362195014953613, "learning_rate": 7.753592260634514e-06, "loss": 0.0696, "step": 4720 }, { "epoch": 6.056338028169014, "grad_norm": 6.469157695770264, "learning_rate": 7.746478873239436e-06, "loss": 0.0656, "step": 4730 }, { "epoch": 6.069142125480154, "grad_norm": 8.421032905578613, "learning_rate": 7.73936548584436e-06, "loss": 0.0204, "step": 4740 }, { "epoch": 6.081946222791293, "grad_norm": 7.738709449768066, "learning_rate": 7.732252098449282e-06, "loss": 0.0368, "step": 4750 }, { "epoch": 6.094750320102433, "grad_norm": 5.2488579750061035, "learning_rate": 7.725138711054204e-06, "loss": 0.028, "step": 4760 }, { "epoch": 6.1075544174135725, "grad_norm": 7.755075454711914, "learning_rate": 7.718025323659127e-06, "loss": 0.0503, "step": 4770 }, { "epoch": 6.120358514724712, "grad_norm": 0.0002441200049361214, "learning_rate": 7.71091193626405e-06, "loss": 0.0276, "step": 4780 }, { "epoch": 6.133162612035852, "grad_norm": 0.39838549494743347, "learning_rate": 7.703798548868971e-06, "loss": 0.0304, "step": 4790 }, { "epoch": 6.145966709346991, "grad_norm": 0.024649253115057945, "learning_rate": 7.696685161473895e-06, "loss": 0.0197, "step": 4800 }, { "epoch": 6.158770806658131, "grad_norm": 1.9696086645126343, "learning_rate": 7.689571774078817e-06, "loss": 0.0203, "step": 4810 }, { "epoch": 6.17157490396927, "grad_norm": 10.748801231384277, "learning_rate": 7.68245838668374e-06, "loss": 0.0207, "step": 4820 }, { "epoch": 6.184379001280409, "grad_norm": 11.485845565795898, "learning_rate": 7.675344999288662e-06, "loss": 0.0256, "step": 4830 }, { "epoch": 6.197183098591549, "grad_norm": 7.258970260620117, "learning_rate": 7.668231611893584e-06, "loss": 0.0243, "step": 4840 }, { "epoch": 6.209987195902689, "grad_norm": 10.128866195678711, "learning_rate": 7.661118224498508e-06, "loss": 0.0256, "step": 4850 }, { "epoch": 6.222791293213828, "grad_norm": 0.12086392194032669, "learning_rate": 7.65400483710343e-06, "loss": 0.021, "step": 4860 }, { "epoch": 6.235595390524968, "grad_norm": 12.103080749511719, "learning_rate": 7.646891449708352e-06, "loss": 0.0452, "step": 4870 }, { "epoch": 6.2483994878361075, "grad_norm": 8.8416166305542, "learning_rate": 7.639778062313274e-06, "loss": 0.0325, "step": 4880 }, { "epoch": 6.261203585147247, "grad_norm": 3.6297500133514404, "learning_rate": 7.632664674918196e-06, "loss": 0.0213, "step": 4890 }, { "epoch": 6.274007682458387, "grad_norm": 0.017991766333580017, "learning_rate": 7.625551287523119e-06, "loss": 0.035, "step": 4900 }, { "epoch": 6.286811779769526, "grad_norm": 7.022460460662842, "learning_rate": 7.618437900128041e-06, "loss": 0.0359, "step": 4910 }, { "epoch": 6.299615877080666, "grad_norm": 12.322715759277344, "learning_rate": 7.611324512732964e-06, "loss": 0.0238, "step": 4920 }, { "epoch": 6.312419974391806, "grad_norm": 4.2045159339904785, "learning_rate": 7.604211125337887e-06, "loss": 0.0566, "step": 4930 }, { "epoch": 6.325224071702945, "grad_norm": 0.00022214287309907377, "learning_rate": 7.597097737942809e-06, "loss": 0.0515, "step": 4940 }, { "epoch": 6.338028169014084, "grad_norm": 15.03963565826416, "learning_rate": 7.589984350547731e-06, "loss": 0.0402, "step": 4950 }, { "epoch": 6.350832266325224, "grad_norm": 6.871021747589111, "learning_rate": 7.5828709631526545e-06, "loss": 0.0206, "step": 4960 }, { "epoch": 6.363636363636363, "grad_norm": 0.12720976769924164, "learning_rate": 7.5757575757575764e-06, "loss": 0.0344, "step": 4970 }, { "epoch": 6.376440460947503, "grad_norm": 13.747876167297363, "learning_rate": 7.568644188362498e-06, "loss": 0.051, "step": 4980 }, { "epoch": 6.389244558258643, "grad_norm": 0.0013262873981148005, "learning_rate": 7.561530800967421e-06, "loss": 0.0246, "step": 4990 }, { "epoch": 6.402048655569782, "grad_norm": 10.661136627197266, "learning_rate": 7.554417413572344e-06, "loss": 0.0507, "step": 5000 }, { "epoch": 6.414852752880922, "grad_norm": 11.273238182067871, "learning_rate": 7.547304026177267e-06, "loss": 0.057, "step": 5010 }, { "epoch": 6.4276568501920615, "grad_norm": 5.872365474700928, "learning_rate": 7.540190638782189e-06, "loss": 0.01, "step": 5020 }, { "epoch": 6.440460947503201, "grad_norm": 2.3220229148864746, "learning_rate": 7.533077251387111e-06, "loss": 0.029, "step": 5030 }, { "epoch": 6.453265044814341, "grad_norm": 7.182269096374512, "learning_rate": 7.525963863992034e-06, "loss": 0.0378, "step": 5040 }, { "epoch": 6.46606914212548, "grad_norm": 10.79238510131836, "learning_rate": 7.518850476596956e-06, "loss": 0.0389, "step": 5050 }, { "epoch": 6.47887323943662, "grad_norm": 6.862414360046387, "learning_rate": 7.511737089201878e-06, "loss": 0.0416, "step": 5060 }, { "epoch": 6.49167733674776, "grad_norm": 5.202556133270264, "learning_rate": 7.504623701806801e-06, "loss": 0.0087, "step": 5070 }, { "epoch": 6.504481434058899, "grad_norm": 21.767629623413086, "learning_rate": 7.497510314411724e-06, "loss": 0.0667, "step": 5080 }, { "epoch": 6.517285531370039, "grad_norm": 0.04317400977015495, "learning_rate": 7.490396927016646e-06, "loss": 0.0331, "step": 5090 }, { "epoch": 6.530089628681178, "grad_norm": 9.36784553527832, "learning_rate": 7.4832835396215686e-06, "loss": 0.0526, "step": 5100 }, { "epoch": 6.542893725992317, "grad_norm": 7.555966377258301, "learning_rate": 7.4761701522264905e-06, "loss": 0.0183, "step": 5110 }, { "epoch": 6.555697823303457, "grad_norm": 0.0007740323198959231, "learning_rate": 7.469056764831414e-06, "loss": 0.0211, "step": 5120 }, { "epoch": 6.5685019206145965, "grad_norm": 3.4399824142456055, "learning_rate": 7.461943377436336e-06, "loss": 0.056, "step": 5130 }, { "epoch": 6.581306017925736, "grad_norm": 10.60245418548584, "learning_rate": 7.454829990041258e-06, "loss": 0.0276, "step": 5140 }, { "epoch": 6.594110115236876, "grad_norm": 4.913111686706543, "learning_rate": 7.447716602646181e-06, "loss": 0.0228, "step": 5150 }, { "epoch": 6.606914212548015, "grad_norm": 2.1112916469573975, "learning_rate": 7.440603215251103e-06, "loss": 0.0212, "step": 5160 }, { "epoch": 6.619718309859155, "grad_norm": 5.81282901763916, "learning_rate": 7.433489827856026e-06, "loss": 0.0292, "step": 5170 }, { "epoch": 6.632522407170295, "grad_norm": 0.24074219167232513, "learning_rate": 7.426376440460948e-06, "loss": 0.0304, "step": 5180 }, { "epoch": 6.645326504481434, "grad_norm": 12.57872200012207, "learning_rate": 7.41926305306587e-06, "loss": 0.0585, "step": 5190 }, { "epoch": 6.658130601792574, "grad_norm": 0.13878846168518066, "learning_rate": 7.412149665670792e-06, "loss": 0.0512, "step": 5200 }, { "epoch": 6.6709346991037135, "grad_norm": 0.16379255056381226, "learning_rate": 7.405036278275716e-06, "loss": 0.0046, "step": 5210 }, { "epoch": 6.683738796414853, "grad_norm": 1.4986361265182495, "learning_rate": 7.397922890880638e-06, "loss": 0.0328, "step": 5220 }, { "epoch": 6.696542893725992, "grad_norm": 4.881646156311035, "learning_rate": 7.390809503485561e-06, "loss": 0.0341, "step": 5230 }, { "epoch": 6.7093469910371315, "grad_norm": 0.001330950646661222, "learning_rate": 7.383696116090483e-06, "loss": 0.0134, "step": 5240 }, { "epoch": 6.722151088348271, "grad_norm": 3.7451059818267822, "learning_rate": 7.376582728695405e-06, "loss": 0.004, "step": 5250 }, { "epoch": 6.734955185659411, "grad_norm": 0.7643628716468811, "learning_rate": 7.369469341300328e-06, "loss": 0.0304, "step": 5260 }, { "epoch": 6.74775928297055, "grad_norm": 0.0019298652186989784, "learning_rate": 7.36235595390525e-06, "loss": 0.05, "step": 5270 }, { "epoch": 6.76056338028169, "grad_norm": 3.20302414894104, "learning_rate": 7.355242566510172e-06, "loss": 0.0342, "step": 5280 }, { "epoch": 6.77336747759283, "grad_norm": 0.13230662047863007, "learning_rate": 7.348129179115096e-06, "loss": 0.0113, "step": 5290 }, { "epoch": 6.786171574903969, "grad_norm": 5.334699630737305, "learning_rate": 7.341015791720018e-06, "loss": 0.0184, "step": 5300 }, { "epoch": 6.798975672215109, "grad_norm": 12.816052436828613, "learning_rate": 7.33390240432494e-06, "loss": 0.0157, "step": 5310 }, { "epoch": 6.8117797695262485, "grad_norm": 0.0011945842998102307, "learning_rate": 7.3267890169298625e-06, "loss": 0.0271, "step": 5320 }, { "epoch": 6.824583866837388, "grad_norm": 0.00851504784077406, "learning_rate": 7.319675629534785e-06, "loss": 0.0081, "step": 5330 }, { "epoch": 6.837387964148528, "grad_norm": 8.16347885131836, "learning_rate": 7.312562242139708e-06, "loss": 0.0297, "step": 5340 }, { "epoch": 6.850192061459667, "grad_norm": 0.011742496863007545, "learning_rate": 7.30544885474463e-06, "loss": 0.0421, "step": 5350 }, { "epoch": 6.862996158770807, "grad_norm": 0.003784577129408717, "learning_rate": 7.298335467349552e-06, "loss": 0.0099, "step": 5360 }, { "epoch": 6.875800256081947, "grad_norm": 14.078619003295898, "learning_rate": 7.291222079954476e-06, "loss": 0.0531, "step": 5370 }, { "epoch": 6.888604353393085, "grad_norm": 0.021575234830379486, "learning_rate": 7.2841086925593975e-06, "loss": 0.0118, "step": 5380 }, { "epoch": 6.901408450704225, "grad_norm": 8.078088760375977, "learning_rate": 7.2769953051643195e-06, "loss": 0.047, "step": 5390 }, { "epoch": 6.914212548015365, "grad_norm": 1.0379023551940918, "learning_rate": 7.269881917769242e-06, "loss": 0.0448, "step": 5400 }, { "epoch": 6.927016645326504, "grad_norm": 5.065972328186035, "learning_rate": 7.262768530374164e-06, "loss": 0.0775, "step": 5410 }, { "epoch": 6.939820742637644, "grad_norm": 0.5640221238136292, "learning_rate": 7.255655142979087e-06, "loss": 0.0343, "step": 5420 }, { "epoch": 6.9526248399487836, "grad_norm": 3.0227553844451904, "learning_rate": 7.24854175558401e-06, "loss": 0.0505, "step": 5430 }, { "epoch": 6.965428937259923, "grad_norm": 0.05710272863507271, "learning_rate": 7.241428368188932e-06, "loss": 0.0235, "step": 5440 }, { "epoch": 6.978233034571063, "grad_norm": 0.04587353393435478, "learning_rate": 7.234314980793855e-06, "loss": 0.0324, "step": 5450 }, { "epoch": 6.991037131882202, "grad_norm": 2.7700085639953613, "learning_rate": 7.227201593398777e-06, "loss": 0.0083, "step": 5460 }, { "epoch": 7.003841229193342, "grad_norm": 1.2235318536113482e-05, "learning_rate": 7.220088206003699e-06, "loss": 0.0151, "step": 5470 }, { "epoch": 7.016645326504482, "grad_norm": 0.02161538228392601, "learning_rate": 7.212974818608622e-06, "loss": 0.0093, "step": 5480 }, { "epoch": 7.029449423815621, "grad_norm": 2.8628182411193848, "learning_rate": 7.205861431213544e-06, "loss": 0.0059, "step": 5490 }, { "epoch": 7.042253521126761, "grad_norm": 0.0012376749655231833, "learning_rate": 7.198748043818467e-06, "loss": 0.0713, "step": 5500 }, { "epoch": 7.0550576184379, "grad_norm": 4.976552486419678, "learning_rate": 7.19163465642339e-06, "loss": 0.0377, "step": 5510 }, { "epoch": 7.067861715749039, "grad_norm": 8.837820053100586, "learning_rate": 7.184521269028312e-06, "loss": 0.0432, "step": 5520 }, { "epoch": 7.080665813060179, "grad_norm": 2.703232765197754, "learning_rate": 7.177407881633235e-06, "loss": 0.0416, "step": 5530 }, { "epoch": 7.093469910371319, "grad_norm": 1.4878956079483032, "learning_rate": 7.170294494238157e-06, "loss": 0.0112, "step": 5540 }, { "epoch": 7.106274007682458, "grad_norm": 4.473015785217285, "learning_rate": 7.163181106843079e-06, "loss": 0.0544, "step": 5550 }, { "epoch": 7.119078104993598, "grad_norm": 0.00038376348675228655, "learning_rate": 7.156067719448002e-06, "loss": 0.0099, "step": 5560 }, { "epoch": 7.1318822023047375, "grad_norm": 0.18693281710147858, "learning_rate": 7.148954332052924e-06, "loss": 0.0392, "step": 5570 }, { "epoch": 7.144686299615877, "grad_norm": 6.844393253326416, "learning_rate": 7.141840944657847e-06, "loss": 0.0567, "step": 5580 }, { "epoch": 7.157490396927017, "grad_norm": 0.0002584387257229537, "learning_rate": 7.1347275572627695e-06, "loss": 0.0398, "step": 5590 }, { "epoch": 7.170294494238156, "grad_norm": 0.00515276612713933, "learning_rate": 7.1276141698676914e-06, "loss": 0.0573, "step": 5600 }, { "epoch": 7.183098591549296, "grad_norm": 0.00010981922969222069, "learning_rate": 7.120500782472613e-06, "loss": 0.0279, "step": 5610 }, { "epoch": 7.195902688860436, "grad_norm": 2.1620774269104004, "learning_rate": 7.113387395077537e-06, "loss": 0.0387, "step": 5620 }, { "epoch": 7.208706786171575, "grad_norm": 0.3693879246711731, "learning_rate": 7.106274007682459e-06, "loss": 0.0131, "step": 5630 }, { "epoch": 7.221510883482715, "grad_norm": 1.0489726066589355, "learning_rate": 7.099160620287382e-06, "loss": 0.0208, "step": 5640 }, { "epoch": 7.2343149807938545, "grad_norm": 6.87832498550415, "learning_rate": 7.092047232892304e-06, "loss": 0.0395, "step": 5650 }, { "epoch": 7.247119078104993, "grad_norm": 7.118307113647461, "learning_rate": 7.084933845497226e-06, "loss": 0.0184, "step": 5660 }, { "epoch": 7.259923175416133, "grad_norm": 9.100848197937012, "learning_rate": 7.077820458102149e-06, "loss": 0.0553, "step": 5670 }, { "epoch": 7.2727272727272725, "grad_norm": 5.670505523681641, "learning_rate": 7.070707070707071e-06, "loss": 0.033, "step": 5680 }, { "epoch": 7.285531370038412, "grad_norm": 2.2115561962127686, "learning_rate": 7.063593683311993e-06, "loss": 0.0217, "step": 5690 }, { "epoch": 7.298335467349552, "grad_norm": 2.2103259563446045, "learning_rate": 7.056480295916917e-06, "loss": 0.0498, "step": 5700 }, { "epoch": 7.311139564660691, "grad_norm": 8.725231170654297, "learning_rate": 7.049366908521839e-06, "loss": 0.0306, "step": 5710 }, { "epoch": 7.323943661971831, "grad_norm": 13.23337173461914, "learning_rate": 7.042253521126761e-06, "loss": 0.0217, "step": 5720 }, { "epoch": 7.336747759282971, "grad_norm": 0.0009325257269665599, "learning_rate": 7.0351401337316836e-06, "loss": 0.0401, "step": 5730 }, { "epoch": 7.34955185659411, "grad_norm": 0.1379578411579132, "learning_rate": 7.0280267463366055e-06, "loss": 0.0294, "step": 5740 }, { "epoch": 7.36235595390525, "grad_norm": 0.005301289260387421, "learning_rate": 7.020913358941529e-06, "loss": 0.0362, "step": 5750 }, { "epoch": 7.3751600512163895, "grad_norm": 9.648992538452148, "learning_rate": 7.013799971546451e-06, "loss": 0.0201, "step": 5760 }, { "epoch": 7.387964148527529, "grad_norm": 0.0003817138494923711, "learning_rate": 7.006686584151373e-06, "loss": 0.0446, "step": 5770 }, { "epoch": 7.400768245838669, "grad_norm": 0.9943172335624695, "learning_rate": 6.999573196756297e-06, "loss": 0.004, "step": 5780 }, { "epoch": 7.4135723431498075, "grad_norm": 6.500863075256348, "learning_rate": 6.992459809361219e-06, "loss": 0.0093, "step": 5790 }, { "epoch": 7.426376440460947, "grad_norm": 11.811480522155762, "learning_rate": 6.985346421966141e-06, "loss": 0.049, "step": 5800 }, { "epoch": 7.439180537772087, "grad_norm": 1.2716519832611084, "learning_rate": 6.978233034571063e-06, "loss": 0.0103, "step": 5810 }, { "epoch": 7.451984635083226, "grad_norm": 1.4747167825698853, "learning_rate": 6.971119647175985e-06, "loss": 0.013, "step": 5820 }, { "epoch": 7.464788732394366, "grad_norm": 8.465987205505371, "learning_rate": 6.964006259780907e-06, "loss": 0.0459, "step": 5830 }, { "epoch": 7.477592829705506, "grad_norm": 5.97678804397583, "learning_rate": 6.956892872385831e-06, "loss": 0.0396, "step": 5840 }, { "epoch": 7.490396927016645, "grad_norm": 5.852774143218994, "learning_rate": 6.949779484990753e-06, "loss": 0.0431, "step": 5850 }, { "epoch": 7.503201024327785, "grad_norm": 4.193504810333252, "learning_rate": 6.9426660975956765e-06, "loss": 0.015, "step": 5860 }, { "epoch": 7.5160051216389245, "grad_norm": 0.0008560824790038168, "learning_rate": 6.9355527102005985e-06, "loss": 0.0185, "step": 5870 }, { "epoch": 7.528809218950064, "grad_norm": 10.773675918579102, "learning_rate": 6.9284393228055204e-06, "loss": 0.0394, "step": 5880 }, { "epoch": 7.541613316261204, "grad_norm": 0.01661304011940956, "learning_rate": 6.921325935410443e-06, "loss": 0.0216, "step": 5890 }, { "epoch": 7.554417413572343, "grad_norm": 0.007579338736832142, "learning_rate": 6.914212548015365e-06, "loss": 0.0421, "step": 5900 }, { "epoch": 7.567221510883483, "grad_norm": 2.102788209915161, "learning_rate": 6.907099160620287e-06, "loss": 0.039, "step": 5910 }, { "epoch": 7.580025608194623, "grad_norm": 0.01709793694317341, "learning_rate": 6.899985773225211e-06, "loss": 0.0278, "step": 5920 }, { "epoch": 7.592829705505762, "grad_norm": 5.662108421325684, "learning_rate": 6.892872385830133e-06, "loss": 0.0185, "step": 5930 }, { "epoch": 7.605633802816901, "grad_norm": 2.4857561588287354, "learning_rate": 6.885758998435055e-06, "loss": 0.0244, "step": 5940 }, { "epoch": 7.618437900128041, "grad_norm": 0.0007264100131578743, "learning_rate": 6.878645611039978e-06, "loss": 0.0282, "step": 5950 }, { "epoch": 7.63124199743918, "grad_norm": 2.5964035987854004, "learning_rate": 6.8715322236449e-06, "loss": 0.0225, "step": 5960 }, { "epoch": 7.64404609475032, "grad_norm": 13.879901885986328, "learning_rate": 6.864418836249823e-06, "loss": 0.0319, "step": 5970 }, { "epoch": 7.65685019206146, "grad_norm": 11.44944953918457, "learning_rate": 6.857305448854745e-06, "loss": 0.0355, "step": 5980 }, { "epoch": 7.669654289372599, "grad_norm": 0.35753241181373596, "learning_rate": 6.850192061459667e-06, "loss": 0.0482, "step": 5990 }, { "epoch": 7.682458386683739, "grad_norm": 7.241912841796875, "learning_rate": 6.843078674064591e-06, "loss": 0.0438, "step": 6000 }, { "epoch": 7.6952624839948784, "grad_norm": 13.230270385742188, "learning_rate": 6.8359652866695126e-06, "loss": 0.0334, "step": 6010 }, { "epoch": 7.708066581306018, "grad_norm": 6.799299240112305, "learning_rate": 6.8288518992744345e-06, "loss": 0.0652, "step": 6020 }, { "epoch": 7.720870678617158, "grad_norm": 0.7865540385246277, "learning_rate": 6.821738511879358e-06, "loss": 0.0122, "step": 6030 }, { "epoch": 7.733674775928297, "grad_norm": 0.0007873768918216228, "learning_rate": 6.81462512448428e-06, "loss": 0.0241, "step": 6040 }, { "epoch": 7.746478873239437, "grad_norm": 4.141044616699219, "learning_rate": 6.807511737089203e-06, "loss": 0.0339, "step": 6050 }, { "epoch": 7.759282970550576, "grad_norm": 7.815735340118408, "learning_rate": 6.800398349694125e-06, "loss": 0.0436, "step": 6060 }, { "epoch": 7.772087067861715, "grad_norm": 0.00042812147876247764, "learning_rate": 6.793284962299047e-06, "loss": 0.0353, "step": 6070 }, { "epoch": 7.784891165172855, "grad_norm": 0.0009545993525534868, "learning_rate": 6.7861715749039704e-06, "loss": 0.0269, "step": 6080 }, { "epoch": 7.797695262483995, "grad_norm": 0.08257643133401871, "learning_rate": 6.779058187508892e-06, "loss": 0.0063, "step": 6090 }, { "epoch": 7.810499359795134, "grad_norm": 0.8201851844787598, "learning_rate": 6.771944800113814e-06, "loss": 0.0177, "step": 6100 }, { "epoch": 7.823303457106274, "grad_norm": 6.194039344787598, "learning_rate": 6.764831412718737e-06, "loss": 0.057, "step": 6110 }, { "epoch": 7.8361075544174135, "grad_norm": 16.03792953491211, "learning_rate": 6.75771802532366e-06, "loss": 0.0412, "step": 6120 }, { "epoch": 7.848911651728553, "grad_norm": 0.1203761100769043, "learning_rate": 6.750604637928582e-06, "loss": 0.0185, "step": 6130 }, { "epoch": 7.861715749039693, "grad_norm": 0.013526562601327896, "learning_rate": 6.743491250533505e-06, "loss": 0.0203, "step": 6140 }, { "epoch": 7.874519846350832, "grad_norm": 0.5092633962631226, "learning_rate": 6.736377863138427e-06, "loss": 0.0152, "step": 6150 }, { "epoch": 7.887323943661972, "grad_norm": 1.3552926778793335, "learning_rate": 6.72926447574335e-06, "loss": 0.0052, "step": 6160 }, { "epoch": 7.900128040973112, "grad_norm": 10.61555290222168, "learning_rate": 6.722151088348272e-06, "loss": 0.0347, "step": 6170 }, { "epoch": 7.912932138284251, "grad_norm": 21.161867141723633, "learning_rate": 6.715037700953194e-06, "loss": 0.0255, "step": 6180 }, { "epoch": 7.925736235595391, "grad_norm": 8.365350723266602, "learning_rate": 6.707924313558117e-06, "loss": 0.039, "step": 6190 }, { "epoch": 7.9385403329065305, "grad_norm": 0.0001548419240862131, "learning_rate": 6.70081092616304e-06, "loss": 0.0098, "step": 6200 }, { "epoch": 7.95134443021767, "grad_norm": 1.4583344636776019e-05, "learning_rate": 6.693697538767962e-06, "loss": 0.0044, "step": 6210 }, { "epoch": 7.964148527528809, "grad_norm": 7.527254581451416, "learning_rate": 6.6865841513728845e-06, "loss": 0.0582, "step": 6220 }, { "epoch": 7.9769526248399485, "grad_norm": 0.61320561170578, "learning_rate": 6.6794707639778065e-06, "loss": 0.0317, "step": 6230 }, { "epoch": 7.989756722151088, "grad_norm": 0.19406023621559143, "learning_rate": 6.672357376582728e-06, "loss": 0.0096, "step": 6240 }, { "epoch": 8.002560819462229, "grad_norm": 11.301575660705566, "learning_rate": 6.665243989187652e-06, "loss": 0.0111, "step": 6250 }, { "epoch": 8.015364916773368, "grad_norm": 2.8440339519875124e-05, "learning_rate": 6.658130601792574e-06, "loss": 0.0388, "step": 6260 }, { "epoch": 8.028169014084508, "grad_norm": 3.551042079925537, "learning_rate": 6.651017214397497e-06, "loss": 0.0242, "step": 6270 }, { "epoch": 8.040973111395646, "grad_norm": 0.00017540018598083407, "learning_rate": 6.64390382700242e-06, "loss": 0.0252, "step": 6280 }, { "epoch": 8.053777208706785, "grad_norm": 0.056313656270504, "learning_rate": 6.6367904396073415e-06, "loss": 0.0201, "step": 6290 }, { "epoch": 8.066581306017925, "grad_norm": 0.5794691443443298, "learning_rate": 6.629677052212264e-06, "loss": 0.0032, "step": 6300 }, { "epoch": 8.079385403329065, "grad_norm": 11.873108863830566, "learning_rate": 6.622563664817186e-06, "loss": 0.0219, "step": 6310 }, { "epoch": 8.092189500640204, "grad_norm": 1.292667031288147, "learning_rate": 6.615450277422108e-06, "loss": 0.0086, "step": 6320 }, { "epoch": 8.104993597951344, "grad_norm": 12.062530517578125, "learning_rate": 6.608336890027032e-06, "loss": 0.0408, "step": 6330 }, { "epoch": 8.117797695262484, "grad_norm": 0.20443128049373627, "learning_rate": 6.601223502631954e-06, "loss": 0.0253, "step": 6340 }, { "epoch": 8.130601792573623, "grad_norm": 0.04760503023862839, "learning_rate": 6.594110115236876e-06, "loss": 0.0164, "step": 6350 }, { "epoch": 8.143405889884763, "grad_norm": 4.753417015075684, "learning_rate": 6.586996727841799e-06, "loss": 0.0203, "step": 6360 }, { "epoch": 8.156209987195902, "grad_norm": 5.232663631439209, "learning_rate": 6.579883340446721e-06, "loss": 0.021, "step": 6370 }, { "epoch": 8.169014084507042, "grad_norm": 0.12114059180021286, "learning_rate": 6.572769953051644e-06, "loss": 0.0304, "step": 6380 }, { "epoch": 8.181818181818182, "grad_norm": 0.003908697050064802, "learning_rate": 6.565656565656566e-06, "loss": 0.0151, "step": 6390 }, { "epoch": 8.194622279129321, "grad_norm": 0.0470287948846817, "learning_rate": 6.558543178261488e-06, "loss": 0.0129, "step": 6400 }, { "epoch": 8.207426376440461, "grad_norm": 1.7126655578613281, "learning_rate": 6.551429790866412e-06, "loss": 0.0039, "step": 6410 }, { "epoch": 8.2202304737516, "grad_norm": 0.2720111012458801, "learning_rate": 6.544316403471334e-06, "loss": 0.0293, "step": 6420 }, { "epoch": 8.23303457106274, "grad_norm": 0.17468300461769104, "learning_rate": 6.537203016076256e-06, "loss": 0.0217, "step": 6430 }, { "epoch": 8.24583866837388, "grad_norm": 7.518794059753418, "learning_rate": 6.530089628681178e-06, "loss": 0.008, "step": 6440 }, { "epoch": 8.25864276568502, "grad_norm": 14.347575187683105, "learning_rate": 6.522976241286101e-06, "loss": 0.0285, "step": 6450 }, { "epoch": 8.271446862996159, "grad_norm": 0.0006101279868744314, "learning_rate": 6.515862853891023e-06, "loss": 0.0187, "step": 6460 }, { "epoch": 8.284250960307299, "grad_norm": 0.00817692931741476, "learning_rate": 6.508749466495946e-06, "loss": 0.0125, "step": 6470 }, { "epoch": 8.297055057618438, "grad_norm": 0.004088731482625008, "learning_rate": 6.501636079100868e-06, "loss": 0.0122, "step": 6480 }, { "epoch": 8.309859154929578, "grad_norm": 6.2541022300720215, "learning_rate": 6.4945226917057916e-06, "loss": 0.0114, "step": 6490 }, { "epoch": 8.322663252240718, "grad_norm": 1.1421973705291748, "learning_rate": 6.4874093043107135e-06, "loss": 0.0203, "step": 6500 }, { "epoch": 8.335467349551857, "grad_norm": 2.169551372528076, "learning_rate": 6.4802959169156355e-06, "loss": 0.0284, "step": 6510 }, { "epoch": 8.348271446862997, "grad_norm": 0.16137059032917023, "learning_rate": 6.473182529520558e-06, "loss": 0.0118, "step": 6520 }, { "epoch": 8.361075544174136, "grad_norm": 0.03140103444457054, "learning_rate": 6.466069142125481e-06, "loss": 0.0339, "step": 6530 }, { "epoch": 8.373879641485276, "grad_norm": 2.4710731506347656, "learning_rate": 6.458955754730403e-06, "loss": 0.0211, "step": 6540 }, { "epoch": 8.386683738796416, "grad_norm": 0.015751808881759644, "learning_rate": 6.451842367335326e-06, "loss": 0.0224, "step": 6550 }, { "epoch": 8.399487836107554, "grad_norm": 2.1411986351013184, "learning_rate": 6.444728979940248e-06, "loss": 0.0097, "step": 6560 }, { "epoch": 8.412291933418693, "grad_norm": 2.9007669581915252e-05, "learning_rate": 6.43761559254517e-06, "loss": 0.01, "step": 6570 }, { "epoch": 8.425096030729833, "grad_norm": 0.04188895225524902, "learning_rate": 6.430502205150093e-06, "loss": 0.0136, "step": 6580 }, { "epoch": 8.437900128040972, "grad_norm": 8.879022598266602, "learning_rate": 6.423388817755015e-06, "loss": 0.0147, "step": 6590 }, { "epoch": 8.450704225352112, "grad_norm": 1.0045573711395264, "learning_rate": 6.416275430359938e-06, "loss": 0.0062, "step": 6600 }, { "epoch": 8.463508322663252, "grad_norm": 5.943238735198975, "learning_rate": 6.40916204296486e-06, "loss": 0.0072, "step": 6610 }, { "epoch": 8.476312419974391, "grad_norm": 4.629339694976807, "learning_rate": 6.402048655569783e-06, "loss": 0.0575, "step": 6620 }, { "epoch": 8.489116517285531, "grad_norm": 0.001093133818358183, "learning_rate": 6.394935268174706e-06, "loss": 0.0139, "step": 6630 }, { "epoch": 8.50192061459667, "grad_norm": 1.0518343515286688e-05, "learning_rate": 6.3878218807796276e-06, "loss": 0.0032, "step": 6640 }, { "epoch": 8.51472471190781, "grad_norm": 6.597747325897217, "learning_rate": 6.3807084933845495e-06, "loss": 0.0294, "step": 6650 }, { "epoch": 8.52752880921895, "grad_norm": 2.3145737648010254, "learning_rate": 6.373595105989473e-06, "loss": 0.0046, "step": 6660 }, { "epoch": 8.54033290653009, "grad_norm": 0.000536081031896174, "learning_rate": 6.366481718594395e-06, "loss": 0.0113, "step": 6670 }, { "epoch": 8.553137003841229, "grad_norm": 0.009426339529454708, "learning_rate": 6.359368331199318e-06, "loss": 0.0297, "step": 6680 }, { "epoch": 8.565941101152369, "grad_norm": 2.1363043785095215, "learning_rate": 6.35225494380424e-06, "loss": 0.0146, "step": 6690 }, { "epoch": 8.578745198463508, "grad_norm": 0.0027642608620226383, "learning_rate": 6.345141556409163e-06, "loss": 0.0393, "step": 6700 }, { "epoch": 8.591549295774648, "grad_norm": 0.0019489210098981857, "learning_rate": 6.3380281690140855e-06, "loss": 0.0284, "step": 6710 }, { "epoch": 8.604353393085788, "grad_norm": 11.977884292602539, "learning_rate": 6.330914781619007e-06, "loss": 0.036, "step": 6720 }, { "epoch": 8.617157490396927, "grad_norm": 5.873406887054443, "learning_rate": 6.323801394223929e-06, "loss": 0.0389, "step": 6730 }, { "epoch": 8.629961587708067, "grad_norm": 0.4539124369621277, "learning_rate": 6.316688006828853e-06, "loss": 0.0152, "step": 6740 }, { "epoch": 8.642765685019207, "grad_norm": 6.1115312576293945, "learning_rate": 6.309574619433775e-06, "loss": 0.0278, "step": 6750 }, { "epoch": 8.655569782330346, "grad_norm": 2.270117875013966e-06, "learning_rate": 6.302461232038697e-06, "loss": 0.0033, "step": 6760 }, { "epoch": 8.668373879641486, "grad_norm": 7.985001564025879, "learning_rate": 6.29534784464362e-06, "loss": 0.0204, "step": 6770 }, { "epoch": 8.681177976952625, "grad_norm": 0.03416327014565468, "learning_rate": 6.2882344572485425e-06, "loss": 0.0045, "step": 6780 }, { "epoch": 8.693982074263765, "grad_norm": 0.5900954604148865, "learning_rate": 6.281121069853465e-06, "loss": 0.0086, "step": 6790 }, { "epoch": 8.706786171574905, "grad_norm": 0.004416467621922493, "learning_rate": 6.274007682458387e-06, "loss": 0.0344, "step": 6800 }, { "epoch": 8.719590268886044, "grad_norm": 2.664764642715454, "learning_rate": 6.266894295063309e-06, "loss": 0.0248, "step": 6810 }, { "epoch": 8.732394366197184, "grad_norm": 5.564209938049316, "learning_rate": 6.259780907668233e-06, "loss": 0.0327, "step": 6820 }, { "epoch": 8.745198463508324, "grad_norm": 0.5846773982048035, "learning_rate": 6.252667520273155e-06, "loss": 0.0327, "step": 6830 }, { "epoch": 8.758002560819463, "grad_norm": 7.859292507171631, "learning_rate": 6.245554132878077e-06, "loss": 0.0416, "step": 6840 }, { "epoch": 8.770806658130601, "grad_norm": 2.4465516617055982e-06, "learning_rate": 6.2384407454829995e-06, "loss": 0.0167, "step": 6850 }, { "epoch": 8.78361075544174, "grad_norm": 5.844322681427002, "learning_rate": 6.2313273580879215e-06, "loss": 0.0255, "step": 6860 }, { "epoch": 8.79641485275288, "grad_norm": 13.128824234008789, "learning_rate": 6.224213970692844e-06, "loss": 0.0361, "step": 6870 }, { "epoch": 8.80921895006402, "grad_norm": 5.178446292877197, "learning_rate": 6.217100583297767e-06, "loss": 0.0439, "step": 6880 }, { "epoch": 8.82202304737516, "grad_norm": 0.004210196901112795, "learning_rate": 6.209987195902689e-06, "loss": 0.0193, "step": 6890 }, { "epoch": 8.8348271446863, "grad_norm": 10.143773078918457, "learning_rate": 6.202873808507613e-06, "loss": 0.0248, "step": 6900 }, { "epoch": 8.847631241997439, "grad_norm": 0.19916215538978577, "learning_rate": 6.195760421112535e-06, "loss": 0.025, "step": 6910 }, { "epoch": 8.860435339308578, "grad_norm": 0.9308204650878906, "learning_rate": 6.1886470337174566e-06, "loss": 0.0092, "step": 6920 }, { "epoch": 8.873239436619718, "grad_norm": 3.2534608840942383, "learning_rate": 6.181533646322379e-06, "loss": 0.024, "step": 6930 }, { "epoch": 8.886043533930858, "grad_norm": 2.1089658737182617, "learning_rate": 6.174420258927301e-06, "loss": 0.0315, "step": 6940 }, { "epoch": 8.898847631241997, "grad_norm": 0.04471014067530632, "learning_rate": 6.167306871532224e-06, "loss": 0.0357, "step": 6950 }, { "epoch": 8.911651728553137, "grad_norm": 4.264136791229248, "learning_rate": 6.160193484137147e-06, "loss": 0.0138, "step": 6960 }, { "epoch": 8.924455825864277, "grad_norm": 1.5806071758270264, "learning_rate": 6.153080096742069e-06, "loss": 0.0225, "step": 6970 }, { "epoch": 8.937259923175416, "grad_norm": 5.23403263092041, "learning_rate": 6.145966709346991e-06, "loss": 0.0239, "step": 6980 }, { "epoch": 8.950064020486556, "grad_norm": 0.20942619442939758, "learning_rate": 6.1388533219519144e-06, "loss": 0.0164, "step": 6990 }, { "epoch": 8.962868117797695, "grad_norm": 8.826539993286133, "learning_rate": 6.131739934556836e-06, "loss": 0.0171, "step": 7000 }, { "epoch": 8.975672215108835, "grad_norm": 1.938127040863037, "learning_rate": 6.124626547161759e-06, "loss": 0.0356, "step": 7010 }, { "epoch": 8.988476312419975, "grad_norm": 0.00025622901739552617, "learning_rate": 6.117513159766681e-06, "loss": 0.0393, "step": 7020 }, { "epoch": 9.001280409731114, "grad_norm": 11.618143081665039, "learning_rate": 6.110399772371604e-06, "loss": 0.0227, "step": 7030 }, { "epoch": 9.014084507042254, "grad_norm": 0.8240607976913452, "learning_rate": 6.103286384976527e-06, "loss": 0.0083, "step": 7040 }, { "epoch": 9.026888604353394, "grad_norm": 4.231604099273682, "learning_rate": 6.096172997581449e-06, "loss": 0.0042, "step": 7050 }, { "epoch": 9.039692701664533, "grad_norm": 0.010406497865915298, "learning_rate": 6.089059610186371e-06, "loss": 0.0205, "step": 7060 }, { "epoch": 9.052496798975673, "grad_norm": 6.202728748321533, "learning_rate": 6.081946222791294e-06, "loss": 0.0365, "step": 7070 }, { "epoch": 9.065300896286812, "grad_norm": 11.739533424377441, "learning_rate": 6.074832835396216e-06, "loss": 0.0248, "step": 7080 }, { "epoch": 9.078104993597952, "grad_norm": 3.568795680999756, "learning_rate": 6.067719448001138e-06, "loss": 0.0124, "step": 7090 }, { "epoch": 9.090909090909092, "grad_norm": 12.654638290405273, "learning_rate": 6.060606060606061e-06, "loss": 0.0148, "step": 7100 }, { "epoch": 9.103713188220231, "grad_norm": 5.410876750946045, "learning_rate": 6.053492673210983e-06, "loss": 0.0187, "step": 7110 }, { "epoch": 9.11651728553137, "grad_norm": 0.11263758689165115, "learning_rate": 6.0463792858159066e-06, "loss": 0.0073, "step": 7120 }, { "epoch": 9.129321382842509, "grad_norm": 0.0015013557858765125, "learning_rate": 6.0392658984208285e-06, "loss": 0.0277, "step": 7130 }, { "epoch": 9.142125480153648, "grad_norm": 0.0004593905759975314, "learning_rate": 6.0321525110257505e-06, "loss": 0.0009, "step": 7140 }, { "epoch": 9.154929577464788, "grad_norm": 0.00045078451512381434, "learning_rate": 6.025039123630674e-06, "loss": 0.0058, "step": 7150 }, { "epoch": 9.167733674775928, "grad_norm": 0.8606227040290833, "learning_rate": 6.017925736235596e-06, "loss": 0.0324, "step": 7160 }, { "epoch": 9.180537772087067, "grad_norm": 0.1209554523229599, "learning_rate": 6.010812348840518e-06, "loss": 0.0448, "step": 7170 }, { "epoch": 9.193341869398207, "grad_norm": 0.18431101739406586, "learning_rate": 6.003698961445441e-06, "loss": 0.0091, "step": 7180 }, { "epoch": 9.206145966709347, "grad_norm": 0.00010758891585282981, "learning_rate": 5.996585574050363e-06, "loss": 0.0199, "step": 7190 }, { "epoch": 9.218950064020486, "grad_norm": 0.003244781633839011, "learning_rate": 5.989472186655286e-06, "loss": 0.0252, "step": 7200 }, { "epoch": 9.231754161331626, "grad_norm": 0.028378801420331, "learning_rate": 5.982358799260208e-06, "loss": 0.0103, "step": 7210 }, { "epoch": 9.244558258642765, "grad_norm": 0.2869134545326233, "learning_rate": 5.97524541186513e-06, "loss": 0.0134, "step": 7220 }, { "epoch": 9.257362355953905, "grad_norm": 0.0006323495181277394, "learning_rate": 5.968132024470054e-06, "loss": 0.0229, "step": 7230 }, { "epoch": 9.270166453265045, "grad_norm": 0.1945762187242508, "learning_rate": 5.961018637074976e-06, "loss": 0.0248, "step": 7240 }, { "epoch": 9.282970550576184, "grad_norm": 13.728672981262207, "learning_rate": 5.953905249679898e-06, "loss": 0.0219, "step": 7250 }, { "epoch": 9.295774647887324, "grad_norm": 8.00424575805664, "learning_rate": 5.946791862284821e-06, "loss": 0.022, "step": 7260 }, { "epoch": 9.308578745198464, "grad_norm": 15.362007141113281, "learning_rate": 5.939678474889743e-06, "loss": 0.0247, "step": 7270 }, { "epoch": 9.321382842509603, "grad_norm": 4.30637264251709, "learning_rate": 5.932565087494665e-06, "loss": 0.0105, "step": 7280 }, { "epoch": 9.334186939820743, "grad_norm": 2.203664541244507, "learning_rate": 5.925451700099588e-06, "loss": 0.0187, "step": 7290 }, { "epoch": 9.346991037131883, "grad_norm": 0.19343163073062897, "learning_rate": 5.91833831270451e-06, "loss": 0.0122, "step": 7300 }, { "epoch": 9.359795134443022, "grad_norm": 5.557631015777588, "learning_rate": 5.911224925309434e-06, "loss": 0.0422, "step": 7310 }, { "epoch": 9.372599231754162, "grad_norm": 8.340714454650879, "learning_rate": 5.904111537914356e-06, "loss": 0.0305, "step": 7320 }, { "epoch": 9.385403329065301, "grad_norm": 12.529728889465332, "learning_rate": 5.896998150519278e-06, "loss": 0.0192, "step": 7330 }, { "epoch": 9.398207426376441, "grad_norm": 1.107229232788086, "learning_rate": 5.8898847631242005e-06, "loss": 0.0166, "step": 7340 }, { "epoch": 9.41101152368758, "grad_norm": 9.787753105163574, "learning_rate": 5.882771375729122e-06, "loss": 0.0839, "step": 7350 }, { "epoch": 9.42381562099872, "grad_norm": 11.11371898651123, "learning_rate": 5.875657988334044e-06, "loss": 0.0117, "step": 7360 }, { "epoch": 9.43661971830986, "grad_norm": 11.242392539978027, "learning_rate": 5.868544600938968e-06, "loss": 0.0209, "step": 7370 }, { "epoch": 9.449423815621, "grad_norm": 1.2927099466323853, "learning_rate": 5.86143121354389e-06, "loss": 0.0059, "step": 7380 }, { "epoch": 9.46222791293214, "grad_norm": 0.0002863346308004111, "learning_rate": 5.854317826148812e-06, "loss": 0.0193, "step": 7390 }, { "epoch": 9.475032010243277, "grad_norm": 0.008841932751238346, "learning_rate": 5.8472044387537356e-06, "loss": 0.0468, "step": 7400 }, { "epoch": 9.487836107554417, "grad_norm": 0.0003540738543961197, "learning_rate": 5.8400910513586575e-06, "loss": 0.0333, "step": 7410 }, { "epoch": 9.500640204865556, "grad_norm": 6.120032787322998, "learning_rate": 5.83297766396358e-06, "loss": 0.0244, "step": 7420 }, { "epoch": 9.513444302176696, "grad_norm": 7.007648944854736, "learning_rate": 5.825864276568502e-06, "loss": 0.025, "step": 7430 }, { "epoch": 9.526248399487836, "grad_norm": 2.805581569671631, "learning_rate": 5.818750889173424e-06, "loss": 0.0066, "step": 7440 }, { "epoch": 9.539052496798975, "grad_norm": 0.37064647674560547, "learning_rate": 5.811637501778348e-06, "loss": 0.0421, "step": 7450 }, { "epoch": 9.551856594110115, "grad_norm": 0.0743195042014122, "learning_rate": 5.80452411438327e-06, "loss": 0.0094, "step": 7460 }, { "epoch": 9.564660691421254, "grad_norm": 0.38670673966407776, "learning_rate": 5.797410726988192e-06, "loss": 0.0178, "step": 7470 }, { "epoch": 9.577464788732394, "grad_norm": 0.0007573810289613903, "learning_rate": 5.790297339593115e-06, "loss": 0.0197, "step": 7480 }, { "epoch": 9.590268886043534, "grad_norm": 0.0007631367188878357, "learning_rate": 5.783183952198037e-06, "loss": 0.024, "step": 7490 }, { "epoch": 9.603072983354673, "grad_norm": 4.409999847412109, "learning_rate": 5.776070564802959e-06, "loss": 0.0157, "step": 7500 }, { "epoch": 9.615877080665813, "grad_norm": 0.12372616678476334, "learning_rate": 5.768957177407882e-06, "loss": 0.017, "step": 7510 }, { "epoch": 9.628681177976953, "grad_norm": 4.666311740875244, "learning_rate": 5.761843790012804e-06, "loss": 0.0088, "step": 7520 }, { "epoch": 9.641485275288092, "grad_norm": 0.6963228583335876, "learning_rate": 5.754730402617728e-06, "loss": 0.0112, "step": 7530 }, { "epoch": 9.654289372599232, "grad_norm": 2.7091293334960938, "learning_rate": 5.74761701522265e-06, "loss": 0.0325, "step": 7540 }, { "epoch": 9.667093469910371, "grad_norm": 0.0012502436293289065, "learning_rate": 5.740503627827572e-06, "loss": 0.0288, "step": 7550 }, { "epoch": 9.679897567221511, "grad_norm": 0.4340243339538574, "learning_rate": 5.733390240432495e-06, "loss": 0.0142, "step": 7560 }, { "epoch": 9.69270166453265, "grad_norm": 6.165146827697754, "learning_rate": 5.726276853037417e-06, "loss": 0.0225, "step": 7570 }, { "epoch": 9.70550576184379, "grad_norm": 0.4861792325973511, "learning_rate": 5.719163465642339e-06, "loss": 0.0213, "step": 7580 }, { "epoch": 9.71830985915493, "grad_norm": 0.0006121151964180171, "learning_rate": 5.712050078247262e-06, "loss": 0.0052, "step": 7590 }, { "epoch": 9.73111395646607, "grad_norm": 2.431206703186035, "learning_rate": 5.704936690852184e-06, "loss": 0.0336, "step": 7600 }, { "epoch": 9.74391805377721, "grad_norm": 0.45705634355545044, "learning_rate": 5.697823303457106e-06, "loss": 0.0145, "step": 7610 }, { "epoch": 9.756722151088349, "grad_norm": 0.5682559609413147, "learning_rate": 5.6907099160620295e-06, "loss": 0.0196, "step": 7620 }, { "epoch": 9.769526248399488, "grad_norm": 6.153069019317627, "learning_rate": 5.683596528666951e-06, "loss": 0.0156, "step": 7630 }, { "epoch": 9.782330345710628, "grad_norm": 5.149827480316162, "learning_rate": 5.676483141271874e-06, "loss": 0.015, "step": 7640 }, { "epoch": 9.795134443021768, "grad_norm": 0.10265146195888519, "learning_rate": 5.669369753876797e-06, "loss": 0.0185, "step": 7650 }, { "epoch": 9.807938540332907, "grad_norm": 0.0014299021568149328, "learning_rate": 5.662256366481719e-06, "loss": 0.0103, "step": 7660 }, { "epoch": 9.820742637644045, "grad_norm": 0.0033280113711953163, "learning_rate": 5.655142979086642e-06, "loss": 0.0267, "step": 7670 }, { "epoch": 9.833546734955185, "grad_norm": 2.4432367354165763e-05, "learning_rate": 5.648029591691564e-06, "loss": 0.0006, "step": 7680 }, { "epoch": 9.846350832266324, "grad_norm": 0.46646037697792053, "learning_rate": 5.640916204296486e-06, "loss": 0.0117, "step": 7690 }, { "epoch": 9.859154929577464, "grad_norm": 1.7182458639144897, "learning_rate": 5.633802816901409e-06, "loss": 0.034, "step": 7700 }, { "epoch": 9.871959026888604, "grad_norm": 3.254436492919922, "learning_rate": 5.626689429506331e-06, "loss": 0.0205, "step": 7710 }, { "epoch": 9.884763124199743, "grad_norm": 0.0007287386106327176, "learning_rate": 5.619576042111253e-06, "loss": 0.0165, "step": 7720 }, { "epoch": 9.897567221510883, "grad_norm": 2.78517746925354, "learning_rate": 5.612462654716177e-06, "loss": 0.0181, "step": 7730 }, { "epoch": 9.910371318822023, "grad_norm": 23.15485191345215, "learning_rate": 5.605349267321099e-06, "loss": 0.0297, "step": 7740 }, { "epoch": 9.923175416133162, "grad_norm": 6.08856201171875, "learning_rate": 5.598235879926022e-06, "loss": 0.0222, "step": 7750 }, { "epoch": 9.935979513444302, "grad_norm": 4.3694376945495605, "learning_rate": 5.5911224925309435e-06, "loss": 0.036, "step": 7760 }, { "epoch": 9.948783610755441, "grad_norm": 6.563720703125, "learning_rate": 5.5840091051358655e-06, "loss": 0.0168, "step": 7770 }, { "epoch": 9.961587708066581, "grad_norm": 4.402914524078369, "learning_rate": 5.576895717740789e-06, "loss": 0.025, "step": 7780 }, { "epoch": 9.97439180537772, "grad_norm": 0.16910921037197113, "learning_rate": 5.569782330345711e-06, "loss": 0.0371, "step": 7790 }, { "epoch": 9.98719590268886, "grad_norm": 5.757880210876465, "learning_rate": 5.562668942950633e-06, "loss": 0.0284, "step": 7800 }, { "epoch": 10.0, "grad_norm": 5.801942348480225, "learning_rate": 5.555555555555557e-06, "loss": 0.0189, "step": 7810 }, { "epoch": 10.01280409731114, "grad_norm": 2.259697675704956, "learning_rate": 5.548442168160479e-06, "loss": 0.0199, "step": 7820 }, { "epoch": 10.02560819462228, "grad_norm": 1.5184670686721802, "learning_rate": 5.541328780765401e-06, "loss": 0.0017, "step": 7830 }, { "epoch": 10.038412291933419, "grad_norm": 0.3063870370388031, "learning_rate": 5.534215393370323e-06, "loss": 0.0058, "step": 7840 }, { "epoch": 10.051216389244559, "grad_norm": 0.11822450906038284, "learning_rate": 5.527102005975245e-06, "loss": 0.0282, "step": 7850 }, { "epoch": 10.064020486555698, "grad_norm": 8.766351699829102, "learning_rate": 5.519988618580169e-06, "loss": 0.0527, "step": 7860 }, { "epoch": 10.076824583866838, "grad_norm": 2.0816233158111572, "learning_rate": 5.512875231185091e-06, "loss": 0.029, "step": 7870 }, { "epoch": 10.089628681177977, "grad_norm": 0.004145815037190914, "learning_rate": 5.505761843790013e-06, "loss": 0.0113, "step": 7880 }, { "epoch": 10.102432778489117, "grad_norm": 2.418198347091675, "learning_rate": 5.498648456394936e-06, "loss": 0.014, "step": 7890 }, { "epoch": 10.115236875800257, "grad_norm": 0.01154331024736166, "learning_rate": 5.4915350689998585e-06, "loss": 0.0104, "step": 7900 }, { "epoch": 10.128040973111396, "grad_norm": 0.09246126562356949, "learning_rate": 5.48442168160478e-06, "loss": 0.0136, "step": 7910 }, { "epoch": 10.140845070422536, "grad_norm": 0.03854215145111084, "learning_rate": 5.477308294209703e-06, "loss": 0.0057, "step": 7920 }, { "epoch": 10.153649167733676, "grad_norm": 0.5196567177772522, "learning_rate": 5.470194906814625e-06, "loss": 0.0464, "step": 7930 }, { "epoch": 10.166453265044815, "grad_norm": 0.010794279165565968, "learning_rate": 5.463081519419549e-06, "loss": 0.049, "step": 7940 }, { "epoch": 10.179257362355955, "grad_norm": 0.015262782573699951, "learning_rate": 5.455968132024471e-06, "loss": 0.0374, "step": 7950 }, { "epoch": 10.192061459667093, "grad_norm": 6.748966693878174, "learning_rate": 5.448854744629393e-06, "loss": 0.0483, "step": 7960 }, { "epoch": 10.204865556978232, "grad_norm": 0.933699369430542, "learning_rate": 5.4417413572343155e-06, "loss": 0.0304, "step": 7970 }, { "epoch": 10.217669654289372, "grad_norm": 0.18968996405601501, "learning_rate": 5.434627969839238e-06, "loss": 0.0563, "step": 7980 }, { "epoch": 10.230473751600512, "grad_norm": 0.018745647743344307, "learning_rate": 5.42751458244416e-06, "loss": 0.011, "step": 7990 }, { "epoch": 10.243277848911651, "grad_norm": 0.2195025235414505, "learning_rate": 5.420401195049083e-06, "loss": 0.0092, "step": 8000 }, { "epoch": 10.25608194622279, "grad_norm": 0.0045803626999258995, "learning_rate": 5.413287807654005e-06, "loss": 0.0261, "step": 8010 }, { "epoch": 10.26888604353393, "grad_norm": 7.657270908355713, "learning_rate": 5.406174420258927e-06, "loss": 0.0242, "step": 8020 }, { "epoch": 10.28169014084507, "grad_norm": 0.835312008857727, "learning_rate": 5.3990610328638506e-06, "loss": 0.0274, "step": 8030 }, { "epoch": 10.29449423815621, "grad_norm": 13.976118087768555, "learning_rate": 5.3919476454687725e-06, "loss": 0.0506, "step": 8040 }, { "epoch": 10.30729833546735, "grad_norm": 0.33354663848876953, "learning_rate": 5.384834258073695e-06, "loss": 0.0216, "step": 8050 }, { "epoch": 10.320102432778489, "grad_norm": 0.004234162159264088, "learning_rate": 5.377720870678618e-06, "loss": 0.0047, "step": 8060 }, { "epoch": 10.332906530089629, "grad_norm": 0.058609362691640854, "learning_rate": 5.37060748328354e-06, "loss": 0.0148, "step": 8070 }, { "epoch": 10.345710627400768, "grad_norm": 9.049026489257812, "learning_rate": 5.363494095888463e-06, "loss": 0.0653, "step": 8080 }, { "epoch": 10.358514724711908, "grad_norm": 0.0024408798199146986, "learning_rate": 5.356380708493385e-06, "loss": 0.023, "step": 8090 }, { "epoch": 10.371318822023047, "grad_norm": 4.550828452920541e-05, "learning_rate": 5.349267321098307e-06, "loss": 0.0235, "step": 8100 }, { "epoch": 10.384122919334187, "grad_norm": 6.180349826812744, "learning_rate": 5.34215393370323e-06, "loss": 0.0387, "step": 8110 }, { "epoch": 10.396927016645327, "grad_norm": 0.17973096668720245, "learning_rate": 5.335040546308152e-06, "loss": 0.0169, "step": 8120 }, { "epoch": 10.409731113956466, "grad_norm": 10.550034523010254, "learning_rate": 5.327927158913074e-06, "loss": 0.0294, "step": 8130 }, { "epoch": 10.422535211267606, "grad_norm": 7.681082206545398e-05, "learning_rate": 5.320813771517997e-06, "loss": 0.0168, "step": 8140 }, { "epoch": 10.435339308578746, "grad_norm": 4.540339469909668, "learning_rate": 5.31370038412292e-06, "loss": 0.004, "step": 8150 }, { "epoch": 10.448143405889885, "grad_norm": 0.012163973413407803, "learning_rate": 5.306586996727843e-06, "loss": 0.007, "step": 8160 }, { "epoch": 10.460947503201025, "grad_norm": 12.752285957336426, "learning_rate": 5.299473609332765e-06, "loss": 0.0364, "step": 8170 }, { "epoch": 10.473751600512164, "grad_norm": 6.085719585418701, "learning_rate": 5.292360221937687e-06, "loss": 0.0365, "step": 8180 }, { "epoch": 10.486555697823304, "grad_norm": 0.0005904785939492285, "learning_rate": 5.28524683454261e-06, "loss": 0.0206, "step": 8190 }, { "epoch": 10.499359795134444, "grad_norm": 3.859853313770145e-05, "learning_rate": 5.278133447147532e-06, "loss": 0.0161, "step": 8200 }, { "epoch": 10.512163892445583, "grad_norm": 0.008183574303984642, "learning_rate": 5.271020059752454e-06, "loss": 0.0308, "step": 8210 }, { "epoch": 10.524967989756721, "grad_norm": 10.348427772521973, "learning_rate": 5.263906672357377e-06, "loss": 0.024, "step": 8220 }, { "epoch": 10.53777208706786, "grad_norm": 0.12688617408275604, "learning_rate": 5.2567932849623e-06, "loss": 0.0215, "step": 8230 }, { "epoch": 10.550576184379, "grad_norm": 2.0012238025665283, "learning_rate": 5.249679897567222e-06, "loss": 0.0264, "step": 8240 }, { "epoch": 10.56338028169014, "grad_norm": 0.013163620606064796, "learning_rate": 5.2425665101721445e-06, "loss": 0.0177, "step": 8250 }, { "epoch": 10.57618437900128, "grad_norm": 0.00017364867380820215, "learning_rate": 5.2354531227770664e-06, "loss": 0.0034, "step": 8260 }, { "epoch": 10.58898847631242, "grad_norm": 6.5215935707092285, "learning_rate": 5.22833973538199e-06, "loss": 0.016, "step": 8270 }, { "epoch": 10.601792573623559, "grad_norm": 7.298775672912598, "learning_rate": 5.221226347986912e-06, "loss": 0.0338, "step": 8280 }, { "epoch": 10.614596670934699, "grad_norm": 14.856698989868164, "learning_rate": 5.214112960591834e-06, "loss": 0.0491, "step": 8290 }, { "epoch": 10.627400768245838, "grad_norm": 0.0034104499500244856, "learning_rate": 5.206999573196757e-06, "loss": 0.0399, "step": 8300 }, { "epoch": 10.640204865556978, "grad_norm": 4.574740409851074, "learning_rate": 5.1998861858016796e-06, "loss": 0.0166, "step": 8310 }, { "epoch": 10.653008962868117, "grad_norm": 0.23741105198860168, "learning_rate": 5.1927727984066015e-06, "loss": 0.0114, "step": 8320 }, { "epoch": 10.665813060179257, "grad_norm": 2.919448137283325, "learning_rate": 5.185659411011524e-06, "loss": 0.002, "step": 8330 }, { "epoch": 10.678617157490397, "grad_norm": 0.0015261388616636395, "learning_rate": 5.178546023616446e-06, "loss": 0.0189, "step": 8340 }, { "epoch": 10.691421254801536, "grad_norm": 0.6882287263870239, "learning_rate": 5.171432636221368e-06, "loss": 0.0264, "step": 8350 }, { "epoch": 10.704225352112676, "grad_norm": 0.0005243959603831172, "learning_rate": 5.164319248826292e-06, "loss": 0.0062, "step": 8360 }, { "epoch": 10.717029449423816, "grad_norm": 0.36187997460365295, "learning_rate": 5.157205861431214e-06, "loss": 0.0273, "step": 8370 }, { "epoch": 10.729833546734955, "grad_norm": 9.366905212402344, "learning_rate": 5.150092474036137e-06, "loss": 0.02, "step": 8380 }, { "epoch": 10.742637644046095, "grad_norm": 0.495855450630188, "learning_rate": 5.1429790866410585e-06, "loss": 0.0308, "step": 8390 }, { "epoch": 10.755441741357235, "grad_norm": 11.331965446472168, "learning_rate": 5.135865699245981e-06, "loss": 0.0119, "step": 8400 }, { "epoch": 10.768245838668374, "grad_norm": 8.107277870178223, "learning_rate": 5.128752311850904e-06, "loss": 0.011, "step": 8410 }, { "epoch": 10.781049935979514, "grad_norm": 7.144583702087402, "learning_rate": 5.121638924455826e-06, "loss": 0.0166, "step": 8420 }, { "epoch": 10.793854033290653, "grad_norm": 0.17072844505310059, "learning_rate": 5.114525537060748e-06, "loss": 0.0187, "step": 8430 }, { "epoch": 10.806658130601793, "grad_norm": 0.8634994626045227, "learning_rate": 5.107412149665672e-06, "loss": 0.0105, "step": 8440 }, { "epoch": 10.819462227912933, "grad_norm": 1.7618540525436401, "learning_rate": 5.100298762270594e-06, "loss": 0.0066, "step": 8450 }, { "epoch": 10.832266325224072, "grad_norm": 0.025283854454755783, "learning_rate": 5.0931853748755164e-06, "loss": 0.0157, "step": 8460 }, { "epoch": 10.845070422535212, "grad_norm": 5.791704177856445, "learning_rate": 5.086071987480438e-06, "loss": 0.0485, "step": 8470 }, { "epoch": 10.857874519846352, "grad_norm": 2.1591832637786865, "learning_rate": 5.078958600085361e-06, "loss": 0.0238, "step": 8480 }, { "epoch": 10.870678617157491, "grad_norm": 0.9552658796310425, "learning_rate": 5.071845212690284e-06, "loss": 0.016, "step": 8490 }, { "epoch": 10.88348271446863, "grad_norm": 0.9378514289855957, "learning_rate": 5.064731825295206e-06, "loss": 0.017, "step": 8500 }, { "epoch": 10.89628681177977, "grad_norm": 5.222459094511578e-06, "learning_rate": 5.057618437900128e-06, "loss": 0.0076, "step": 8510 }, { "epoch": 10.909090909090908, "grad_norm": 1.9423551559448242, "learning_rate": 5.0505050505050515e-06, "loss": 0.0099, "step": 8520 }, { "epoch": 10.921895006402048, "grad_norm": 5.069260597229004, "learning_rate": 5.0433916631099735e-06, "loss": 0.0126, "step": 8530 }, { "epoch": 10.934699103713188, "grad_norm": 2.069833278656006, "learning_rate": 5.036278275714895e-06, "loss": 0.024, "step": 8540 }, { "epoch": 10.947503201024327, "grad_norm": 0.05226731672883034, "learning_rate": 5.029164888319818e-06, "loss": 0.0463, "step": 8550 }, { "epoch": 10.960307298335467, "grad_norm": 8.753764152526855, "learning_rate": 5.02205150092474e-06, "loss": 0.0499, "step": 8560 }, { "epoch": 10.973111395646606, "grad_norm": 10.401298522949219, "learning_rate": 5.014938113529664e-06, "loss": 0.0223, "step": 8570 }, { "epoch": 10.985915492957746, "grad_norm": 6.916590213775635, "learning_rate": 5.007824726134586e-06, "loss": 0.0217, "step": 8580 }, { "epoch": 10.998719590268886, "grad_norm": 0.003497628029435873, "learning_rate": 5.000711338739508e-06, "loss": 0.0389, "step": 8590 }, { "epoch": 11.011523687580025, "grad_norm": 0.23923610150814056, "learning_rate": 4.9935979513444305e-06, "loss": 0.0237, "step": 8600 }, { "epoch": 11.024327784891165, "grad_norm": 12.386022567749023, "learning_rate": 4.986484563949353e-06, "loss": 0.0288, "step": 8610 }, { "epoch": 11.037131882202305, "grad_norm": 5.072197437286377, "learning_rate": 4.979371176554276e-06, "loss": 0.0074, "step": 8620 }, { "epoch": 11.049935979513444, "grad_norm": 8.038137435913086, "learning_rate": 4.972257789159198e-06, "loss": 0.0169, "step": 8630 }, { "epoch": 11.062740076824584, "grad_norm": 12.104462623596191, "learning_rate": 4.96514440176412e-06, "loss": 0.0093, "step": 8640 }, { "epoch": 11.075544174135723, "grad_norm": 0.058964360505342484, "learning_rate": 4.958031014369043e-06, "loss": 0.0291, "step": 8650 }, { "epoch": 11.088348271446863, "grad_norm": 5.205768585205078, "learning_rate": 4.950917626973966e-06, "loss": 0.0332, "step": 8660 }, { "epoch": 11.101152368758003, "grad_norm": 0.00029252548119984567, "learning_rate": 4.9438042395788875e-06, "loss": 0.006, "step": 8670 }, { "epoch": 11.113956466069142, "grad_norm": 0.5262041687965393, "learning_rate": 4.93669085218381e-06, "loss": 0.0186, "step": 8680 }, { "epoch": 11.126760563380282, "grad_norm": 9.941957473754883, "learning_rate": 4.929577464788733e-06, "loss": 0.0351, "step": 8690 }, { "epoch": 11.139564660691422, "grad_norm": 0.08207658678293228, "learning_rate": 4.922464077393655e-06, "loss": 0.0103, "step": 8700 }, { "epoch": 11.152368758002561, "grad_norm": 6.0356526374816895, "learning_rate": 4.915350689998578e-06, "loss": 0.0208, "step": 8710 }, { "epoch": 11.1651728553137, "grad_norm": 0.11295334994792938, "learning_rate": 4.9082373026035e-06, "loss": 0.028, "step": 8720 }, { "epoch": 11.17797695262484, "grad_norm": 0.018641943112015724, "learning_rate": 4.901123915208423e-06, "loss": 0.039, "step": 8730 }, { "epoch": 11.19078104993598, "grad_norm": 0.1579485982656479, "learning_rate": 4.8940105278133446e-06, "loss": 0.0132, "step": 8740 }, { "epoch": 11.20358514724712, "grad_norm": 0.05648108199238777, "learning_rate": 4.886897140418267e-06, "loss": 0.0109, "step": 8750 }, { "epoch": 11.21638924455826, "grad_norm": 8.769384384155273, "learning_rate": 4.87978375302319e-06, "loss": 0.0119, "step": 8760 }, { "epoch": 11.229193341869399, "grad_norm": 0.003436629893258214, "learning_rate": 4.872670365628113e-06, "loss": 0.0249, "step": 8770 }, { "epoch": 11.241997439180539, "grad_norm": 11.030573844909668, "learning_rate": 4.865556978233035e-06, "loss": 0.0244, "step": 8780 }, { "epoch": 11.254801536491676, "grad_norm": 0.7862492203712463, "learning_rate": 4.858443590837958e-06, "loss": 0.0105, "step": 8790 }, { "epoch": 11.267605633802816, "grad_norm": 0.029348324984312057, "learning_rate": 4.85133020344288e-06, "loss": 0.007, "step": 8800 }, { "epoch": 11.280409731113956, "grad_norm": 7.6412129402160645, "learning_rate": 4.8442168160478025e-06, "loss": 0.0241, "step": 8810 }, { "epoch": 11.293213828425095, "grad_norm": 2.5081536769866943, "learning_rate": 4.837103428652724e-06, "loss": 0.0102, "step": 8820 }, { "epoch": 11.306017925736235, "grad_norm": 0.30100366473197937, "learning_rate": 4.829990041257647e-06, "loss": 0.0002, "step": 8830 }, { "epoch": 11.318822023047375, "grad_norm": 0.003410755656659603, "learning_rate": 4.82287665386257e-06, "loss": 0.0032, "step": 8840 }, { "epoch": 11.331626120358514, "grad_norm": 0.5186023712158203, "learning_rate": 4.815763266467493e-06, "loss": 0.0174, "step": 8850 }, { "epoch": 11.344430217669654, "grad_norm": 4.904882907867432, "learning_rate": 4.808649879072415e-06, "loss": 0.0087, "step": 8860 }, { "epoch": 11.357234314980793, "grad_norm": 0.03239725902676582, "learning_rate": 4.8015364916773375e-06, "loss": 0.0237, "step": 8870 }, { "epoch": 11.370038412291933, "grad_norm": 0.4754350781440735, "learning_rate": 4.7944231042822595e-06, "loss": 0.016, "step": 8880 }, { "epoch": 11.382842509603073, "grad_norm": 0.16476941108703613, "learning_rate": 4.7873097168871814e-06, "loss": 0.0139, "step": 8890 }, { "epoch": 11.395646606914212, "grad_norm": 14.9800386428833, "learning_rate": 4.780196329492104e-06, "loss": 0.034, "step": 8900 }, { "epoch": 11.408450704225352, "grad_norm": 5.938887119293213, "learning_rate": 4.773082942097027e-06, "loss": 0.0167, "step": 8910 }, { "epoch": 11.421254801536492, "grad_norm": 0.10481268912553787, "learning_rate": 4.76596955470195e-06, "loss": 0.0093, "step": 8920 }, { "epoch": 11.434058898847631, "grad_norm": 5.669556617736816, "learning_rate": 4.758856167306872e-06, "loss": 0.0132, "step": 8930 }, { "epoch": 11.446862996158771, "grad_norm": 6.91036319732666, "learning_rate": 4.751742779911795e-06, "loss": 0.0071, "step": 8940 }, { "epoch": 11.45966709346991, "grad_norm": 4.3748860359191895, "learning_rate": 4.7446293925167165e-06, "loss": 0.0249, "step": 8950 }, { "epoch": 11.47247119078105, "grad_norm": 0.00023807135585229844, "learning_rate": 4.737516005121639e-06, "loss": 0.0434, "step": 8960 }, { "epoch": 11.48527528809219, "grad_norm": 0.00908487569540739, "learning_rate": 4.730402617726561e-06, "loss": 0.0138, "step": 8970 }, { "epoch": 11.49807938540333, "grad_norm": 0.005840748082846403, "learning_rate": 4.723289230331484e-06, "loss": 0.0258, "step": 8980 }, { "epoch": 11.510883482714469, "grad_norm": 0.44342947006225586, "learning_rate": 4.716175842936407e-06, "loss": 0.0103, "step": 8990 }, { "epoch": 11.523687580025609, "grad_norm": 11.988405227661133, "learning_rate": 4.709062455541329e-06, "loss": 0.0327, "step": 9000 }, { "epoch": 11.536491677336748, "grad_norm": 0.008092625066637993, "learning_rate": 4.701949068146252e-06, "loss": 0.0412, "step": 9010 }, { "epoch": 11.549295774647888, "grad_norm": 0.0011768820695579052, "learning_rate": 4.694835680751174e-06, "loss": 0.0085, "step": 9020 }, { "epoch": 11.562099871959028, "grad_norm": 6.319904327392578, "learning_rate": 4.687722293356096e-06, "loss": 0.0094, "step": 9030 }, { "epoch": 11.574903969270167, "grad_norm": 0.2335677444934845, "learning_rate": 4.680608905961019e-06, "loss": 0.0237, "step": 9040 }, { "epoch": 11.587708066581307, "grad_norm": 0.007396308239549398, "learning_rate": 4.673495518565941e-06, "loss": 0.0219, "step": 9050 }, { "epoch": 11.600512163892446, "grad_norm": 0.5111904144287109, "learning_rate": 4.666382131170864e-06, "loss": 0.0287, "step": 9060 }, { "epoch": 11.613316261203586, "grad_norm": 0.300795316696167, "learning_rate": 4.659268743775787e-06, "loss": 0.0084, "step": 9070 }, { "epoch": 11.626120358514724, "grad_norm": 14.654006004333496, "learning_rate": 4.652155356380709e-06, "loss": 0.0211, "step": 9080 }, { "epoch": 11.638924455825864, "grad_norm": 4.150613784790039, "learning_rate": 4.6450419689856314e-06, "loss": 0.0096, "step": 9090 }, { "epoch": 11.651728553137003, "grad_norm": 0.2781789302825928, "learning_rate": 4.637928581590554e-06, "loss": 0.0221, "step": 9100 }, { "epoch": 11.664532650448143, "grad_norm": 4.838508605957031, "learning_rate": 4.630815194195476e-06, "loss": 0.0031, "step": 9110 }, { "epoch": 11.677336747759282, "grad_norm": 0.7836334705352783, "learning_rate": 4.623701806800399e-06, "loss": 0.0158, "step": 9120 }, { "epoch": 11.690140845070422, "grad_norm": 0.0683254599571228, "learning_rate": 4.616588419405321e-06, "loss": 0.0427, "step": 9130 }, { "epoch": 11.702944942381562, "grad_norm": 0.3021542429924011, "learning_rate": 4.609475032010244e-06, "loss": 0.026, "step": 9140 }, { "epoch": 11.715749039692701, "grad_norm": 5.465481281280518, "learning_rate": 4.602361644615166e-06, "loss": 0.0076, "step": 9150 }, { "epoch": 11.728553137003841, "grad_norm": 4.723483085632324, "learning_rate": 4.5952482572200885e-06, "loss": 0.0191, "step": 9160 }, { "epoch": 11.74135723431498, "grad_norm": 0.0921323224902153, "learning_rate": 4.588134869825011e-06, "loss": 0.0097, "step": 9170 }, { "epoch": 11.75416133162612, "grad_norm": 0.001626187120564282, "learning_rate": 4.581021482429934e-06, "loss": 0.0228, "step": 9180 }, { "epoch": 11.76696542893726, "grad_norm": 7.186104774475098, "learning_rate": 4.573908095034856e-06, "loss": 0.0129, "step": 9190 }, { "epoch": 11.7797695262484, "grad_norm": 0.004152664449065924, "learning_rate": 4.566794707639778e-06, "loss": 0.028, "step": 9200 }, { "epoch": 11.792573623559539, "grad_norm": 0.25036826729774475, "learning_rate": 4.559681320244701e-06, "loss": 0.0098, "step": 9210 }, { "epoch": 11.805377720870679, "grad_norm": 12.491260528564453, "learning_rate": 4.5525679328496236e-06, "loss": 0.0292, "step": 9220 }, { "epoch": 11.818181818181818, "grad_norm": 2.6912405490875244, "learning_rate": 4.5454545454545455e-06, "loss": 0.0321, "step": 9230 }, { "epoch": 11.830985915492958, "grad_norm": 3.0300858270493336e-05, "learning_rate": 4.538341158059468e-06, "loss": 0.0019, "step": 9240 }, { "epoch": 11.843790012804098, "grad_norm": 10.405506134033203, "learning_rate": 4.531227770664391e-06, "loss": 0.0104, "step": 9250 }, { "epoch": 11.856594110115237, "grad_norm": 0.45459917187690735, "learning_rate": 4.524114383269313e-06, "loss": 0.0187, "step": 9260 }, { "epoch": 11.869398207426377, "grad_norm": 9.896618843078613, "learning_rate": 4.517000995874236e-06, "loss": 0.0083, "step": 9270 }, { "epoch": 11.882202304737516, "grad_norm": 0.004004519898444414, "learning_rate": 4.509887608479158e-06, "loss": 0.0222, "step": 9280 }, { "epoch": 11.895006402048656, "grad_norm": 0.10266713798046112, "learning_rate": 4.502774221084081e-06, "loss": 0.0118, "step": 9290 }, { "epoch": 11.907810499359796, "grad_norm": 1.3691693544387817, "learning_rate": 4.4956608336890026e-06, "loss": 0.0232, "step": 9300 }, { "epoch": 11.920614596670935, "grad_norm": 0.5837397575378418, "learning_rate": 4.488547446293925e-06, "loss": 0.0055, "step": 9310 }, { "epoch": 11.933418693982075, "grad_norm": 3.9820590019226074, "learning_rate": 4.481434058898848e-06, "loss": 0.0305, "step": 9320 }, { "epoch": 11.946222791293215, "grad_norm": 3.5302162170410156, "learning_rate": 4.474320671503771e-06, "loss": 0.0051, "step": 9330 }, { "epoch": 11.959026888604352, "grad_norm": 0.18159997463226318, "learning_rate": 4.467207284108693e-06, "loss": 0.0204, "step": 9340 }, { "epoch": 11.971830985915492, "grad_norm": 0.0003324251447338611, "learning_rate": 4.460093896713616e-06, "loss": 0.0074, "step": 9350 }, { "epoch": 11.984635083226632, "grad_norm": 23.19733428955078, "learning_rate": 4.452980509318538e-06, "loss": 0.03, "step": 9360 }, { "epoch": 11.997439180537771, "grad_norm": 1.3431735038757324, "learning_rate": 4.4458671219234604e-06, "loss": 0.0143, "step": 9370 }, { "epoch": 12.010243277848911, "grad_norm": 7.359150409698486, "learning_rate": 4.438753734528382e-06, "loss": 0.0078, "step": 9380 }, { "epoch": 12.02304737516005, "grad_norm": 0.020270936191082, "learning_rate": 4.431640347133305e-06, "loss": 0.0154, "step": 9390 }, { "epoch": 12.03585147247119, "grad_norm": 0.002502968767657876, "learning_rate": 4.424526959738228e-06, "loss": 0.0093, "step": 9400 }, { "epoch": 12.04865556978233, "grad_norm": 5.730095386505127, "learning_rate": 4.41741357234315e-06, "loss": 0.0187, "step": 9410 }, { "epoch": 12.06145966709347, "grad_norm": 0.0488029308617115, "learning_rate": 4.410300184948073e-06, "loss": 0.0223, "step": 9420 }, { "epoch": 12.07426376440461, "grad_norm": 0.7696062922477722, "learning_rate": 4.4031867975529955e-06, "loss": 0.0157, "step": 9430 }, { "epoch": 12.087067861715749, "grad_norm": 10.111274719238281, "learning_rate": 4.3960734101579175e-06, "loss": 0.0468, "step": 9440 }, { "epoch": 12.099871959026888, "grad_norm": 0.004602501168847084, "learning_rate": 4.388960022762839e-06, "loss": 0.011, "step": 9450 }, { "epoch": 12.112676056338028, "grad_norm": 0.7055022716522217, "learning_rate": 4.381846635367762e-06, "loss": 0.0226, "step": 9460 }, { "epoch": 12.125480153649168, "grad_norm": 6.4993977546691895, "learning_rate": 4.374733247972685e-06, "loss": 0.0467, "step": 9470 }, { "epoch": 12.138284250960307, "grad_norm": 5.248816967010498, "learning_rate": 4.367619860577608e-06, "loss": 0.0232, "step": 9480 }, { "epoch": 12.151088348271447, "grad_norm": 0.4994298815727234, "learning_rate": 4.36050647318253e-06, "loss": 0.003, "step": 9490 }, { "epoch": 12.163892445582587, "grad_norm": 0.01094064861536026, "learning_rate": 4.3533930857874526e-06, "loss": 0.0212, "step": 9500 }, { "epoch": 12.176696542893726, "grad_norm": 0.0005367989651858807, "learning_rate": 4.346279698392375e-06, "loss": 0.0107, "step": 9510 }, { "epoch": 12.189500640204866, "grad_norm": 6.758755716873566e-06, "learning_rate": 4.339166310997297e-06, "loss": 0.0106, "step": 9520 }, { "epoch": 12.202304737516005, "grad_norm": 0.0007823957712389529, "learning_rate": 4.332052923602219e-06, "loss": 0.0194, "step": 9530 }, { "epoch": 12.215108834827145, "grad_norm": 3.90116548538208, "learning_rate": 4.324939536207142e-06, "loss": 0.0207, "step": 9540 }, { "epoch": 12.227912932138285, "grad_norm": 0.030143318697810173, "learning_rate": 4.317826148812065e-06, "loss": 0.0215, "step": 9550 }, { "epoch": 12.240717029449424, "grad_norm": 0.0011328129330649972, "learning_rate": 4.310712761416987e-06, "loss": 0.0078, "step": 9560 }, { "epoch": 12.253521126760564, "grad_norm": 2.1484808921813965, "learning_rate": 4.30359937402191e-06, "loss": 0.0102, "step": 9570 }, { "epoch": 12.266325224071704, "grad_norm": 7.6401214599609375, "learning_rate": 4.296485986626832e-06, "loss": 0.0315, "step": 9580 }, { "epoch": 12.279129321382843, "grad_norm": 7.893928050994873, "learning_rate": 4.289372599231754e-06, "loss": 0.0173, "step": 9590 }, { "epoch": 12.291933418693983, "grad_norm": 0.3126736879348755, "learning_rate": 4.282259211836677e-06, "loss": 0.0239, "step": 9600 }, { "epoch": 12.304737516005122, "grad_norm": 11.8942289352417, "learning_rate": 4.275145824441599e-06, "loss": 0.0873, "step": 9610 }, { "epoch": 12.317541613316262, "grad_norm": 17.849855422973633, "learning_rate": 4.268032437046522e-06, "loss": 0.0317, "step": 9620 }, { "epoch": 12.3303457106274, "grad_norm": 10.122198104858398, "learning_rate": 4.260919049651444e-06, "loss": 0.0181, "step": 9630 }, { "epoch": 12.34314980793854, "grad_norm": 0.008432975970208645, "learning_rate": 4.253805662256367e-06, "loss": 0.0145, "step": 9640 }, { "epoch": 12.35595390524968, "grad_norm": 0.00012394743680488318, "learning_rate": 4.2466922748612894e-06, "loss": 0.0065, "step": 9650 }, { "epoch": 12.368758002560819, "grad_norm": 6.399505138397217, "learning_rate": 4.239578887466212e-06, "loss": 0.0241, "step": 9660 }, { "epoch": 12.381562099871958, "grad_norm": 0.1277066022157669, "learning_rate": 4.232465500071134e-06, "loss": 0.0237, "step": 9670 }, { "epoch": 12.394366197183098, "grad_norm": 0.00046324028517119586, "learning_rate": 4.225352112676057e-06, "loss": 0.0142, "step": 9680 }, { "epoch": 12.407170294494238, "grad_norm": 0.291615754365921, "learning_rate": 4.218238725280979e-06, "loss": 0.0266, "step": 9690 }, { "epoch": 12.419974391805377, "grad_norm": 7.896489143371582, "learning_rate": 4.211125337885902e-06, "loss": 0.0498, "step": 9700 }, { "epoch": 12.432778489116517, "grad_norm": 9.060142517089844, "learning_rate": 4.204011950490824e-06, "loss": 0.0231, "step": 9710 }, { "epoch": 12.445582586427657, "grad_norm": 1.0932848453521729, "learning_rate": 4.1968985630957465e-06, "loss": 0.0107, "step": 9720 }, { "epoch": 12.458386683738796, "grad_norm": 0.0002862956316675991, "learning_rate": 4.189785175700669e-06, "loss": 0.0323, "step": 9730 }, { "epoch": 12.471190781049936, "grad_norm": 9.223715782165527, "learning_rate": 4.182671788305592e-06, "loss": 0.0317, "step": 9740 }, { "epoch": 12.483994878361075, "grad_norm": 0.011121570132672787, "learning_rate": 4.175558400910514e-06, "loss": 0.0162, "step": 9750 }, { "epoch": 12.496798975672215, "grad_norm": 0.010349276475608349, "learning_rate": 4.168445013515437e-06, "loss": 0.0097, "step": 9760 }, { "epoch": 12.509603072983355, "grad_norm": 0.15885749459266663, "learning_rate": 4.161331626120359e-06, "loss": 0.0071, "step": 9770 }, { "epoch": 12.522407170294494, "grad_norm": 0.07723096013069153, "learning_rate": 4.154218238725281e-06, "loss": 0.0323, "step": 9780 }, { "epoch": 12.535211267605634, "grad_norm": 4.135581016540527, "learning_rate": 4.1471048513302035e-06, "loss": 0.0432, "step": 9790 }, { "epoch": 12.548015364916774, "grad_norm": 0.0004376301367301494, "learning_rate": 4.139991463935126e-06, "loss": 0.0295, "step": 9800 }, { "epoch": 12.560819462227913, "grad_norm": 5.946563720703125, "learning_rate": 4.132878076540049e-06, "loss": 0.0309, "step": 9810 }, { "epoch": 12.573623559539053, "grad_norm": 3.692141056060791, "learning_rate": 4.125764689144971e-06, "loss": 0.0366, "step": 9820 }, { "epoch": 12.586427656850192, "grad_norm": 0.02407948300242424, "learning_rate": 4.118651301749894e-06, "loss": 0.0364, "step": 9830 }, { "epoch": 12.599231754161332, "grad_norm": 0.011938225477933884, "learning_rate": 4.111537914354816e-06, "loss": 0.0127, "step": 9840 }, { "epoch": 12.612035851472472, "grad_norm": 6.503283500671387, "learning_rate": 4.104424526959739e-06, "loss": 0.031, "step": 9850 }, { "epoch": 12.624839948783611, "grad_norm": 0.012061931192874908, "learning_rate": 4.0973111395646605e-06, "loss": 0.0161, "step": 9860 }, { "epoch": 12.637644046094751, "grad_norm": 0.00023004470858722925, "learning_rate": 4.090197752169583e-06, "loss": 0.022, "step": 9870 }, { "epoch": 12.65044814340589, "grad_norm": 0.3356795310974121, "learning_rate": 4.083084364774506e-06, "loss": 0.0322, "step": 9880 }, { "epoch": 12.66325224071703, "grad_norm": 1.0321135520935059, "learning_rate": 4.075970977379428e-06, "loss": 0.0473, "step": 9890 }, { "epoch": 12.676056338028168, "grad_norm": 6.258904933929443, "learning_rate": 4.068857589984351e-06, "loss": 0.0235, "step": 9900 }, { "epoch": 12.688860435339308, "grad_norm": 0.0010816763387992978, "learning_rate": 4.061744202589274e-06, "loss": 0.0246, "step": 9910 }, { "epoch": 12.701664532650447, "grad_norm": 7.38685417175293, "learning_rate": 4.054630815194196e-06, "loss": 0.0115, "step": 9920 }, { "epoch": 12.714468629961587, "grad_norm": 0.09015665203332901, "learning_rate": 4.047517427799118e-06, "loss": 0.0082, "step": 9930 }, { "epoch": 12.727272727272727, "grad_norm": 11.077136993408203, "learning_rate": 4.04040404040404e-06, "loss": 0.0189, "step": 9940 }, { "epoch": 12.740076824583866, "grad_norm": 0.007070339750498533, "learning_rate": 4.033290653008963e-06, "loss": 0.004, "step": 9950 }, { "epoch": 12.752880921895006, "grad_norm": 3.147578001022339, "learning_rate": 4.026177265613886e-06, "loss": 0.0103, "step": 9960 }, { "epoch": 12.765685019206146, "grad_norm": 0.0023142409045249224, "learning_rate": 4.019063878218808e-06, "loss": 0.0177, "step": 9970 }, { "epoch": 12.778489116517285, "grad_norm": 0.024539463222026825, "learning_rate": 4.011950490823731e-06, "loss": 0.0125, "step": 9980 }, { "epoch": 12.791293213828425, "grad_norm": 0.7222532033920288, "learning_rate": 4.0048371034286535e-06, "loss": 0.0123, "step": 9990 }, { "epoch": 12.804097311139564, "grad_norm": 2.2293977963272482e-05, "learning_rate": 3.9977237160335755e-06, "loss": 0.0286, "step": 10000 }, { "epoch": 12.816901408450704, "grad_norm": 0.2896987497806549, "learning_rate": 3.990610328638498e-06, "loss": 0.0313, "step": 10010 }, { "epoch": 12.829705505761844, "grad_norm": 6.5027971267700195, "learning_rate": 3.98349694124342e-06, "loss": 0.0421, "step": 10020 }, { "epoch": 12.842509603072983, "grad_norm": 1.956372618675232, "learning_rate": 3.976383553848343e-06, "loss": 0.0133, "step": 10030 }, { "epoch": 12.855313700384123, "grad_norm": 3.374871253967285, "learning_rate": 3.969270166453265e-06, "loss": 0.0083, "step": 10040 }, { "epoch": 12.868117797695263, "grad_norm": 3.4415929317474365, "learning_rate": 3.962156779058188e-06, "loss": 0.0098, "step": 10050 }, { "epoch": 12.880921895006402, "grad_norm": 17.851837158203125, "learning_rate": 3.9550433916631105e-06, "loss": 0.0248, "step": 10060 }, { "epoch": 12.893725992317542, "grad_norm": 12.793821334838867, "learning_rate": 3.947930004268033e-06, "loss": 0.0073, "step": 10070 }, { "epoch": 12.906530089628681, "grad_norm": 0.11059227585792542, "learning_rate": 3.940816616872955e-06, "loss": 0.027, "step": 10080 }, { "epoch": 12.919334186939821, "grad_norm": 0.3241099715232849, "learning_rate": 3.933703229477877e-06, "loss": 0.0144, "step": 10090 }, { "epoch": 12.93213828425096, "grad_norm": 0.01770693250000477, "learning_rate": 3.9265898420828e-06, "loss": 0.0009, "step": 10100 }, { "epoch": 12.9449423815621, "grad_norm": 0.00023692190006840974, "learning_rate": 3.919476454687723e-06, "loss": 0.0081, "step": 10110 }, { "epoch": 12.95774647887324, "grad_norm": 0.03319963812828064, "learning_rate": 3.912363067292645e-06, "loss": 0.0199, "step": 10120 }, { "epoch": 12.97055057618438, "grad_norm": 0.0035884061362594366, "learning_rate": 3.9052496798975676e-06, "loss": 0.0283, "step": 10130 }, { "epoch": 12.98335467349552, "grad_norm": 0.009372467175126076, "learning_rate": 3.89813629250249e-06, "loss": 0.0034, "step": 10140 }, { "epoch": 12.996158770806659, "grad_norm": 0.025522533804178238, "learning_rate": 3.891022905107412e-06, "loss": 0.027, "step": 10150 }, { "epoch": 13.008962868117798, "grad_norm": 0.03188972547650337, "learning_rate": 3.883909517712335e-06, "loss": 0.0146, "step": 10160 }, { "epoch": 13.021766965428938, "grad_norm": 0.020737428218126297, "learning_rate": 3.876796130317257e-06, "loss": 0.008, "step": 10170 }, { "epoch": 13.034571062740078, "grad_norm": 0.04334618151187897, "learning_rate": 3.86968274292218e-06, "loss": 0.0167, "step": 10180 }, { "epoch": 13.047375160051216, "grad_norm": 2.530782512621954e-05, "learning_rate": 3.862569355527102e-06, "loss": 0.0127, "step": 10190 }, { "epoch": 13.060179257362355, "grad_norm": 1.4986348152160645, "learning_rate": 3.855455968132025e-06, "loss": 0.0378, "step": 10200 }, { "epoch": 13.072983354673495, "grad_norm": 2.3561546186101623e-05, "learning_rate": 3.848342580736947e-06, "loss": 0.0001, "step": 10210 }, { "epoch": 13.085787451984634, "grad_norm": 1.2639870643615723, "learning_rate": 3.84122919334187e-06, "loss": 0.0118, "step": 10220 }, { "epoch": 13.098591549295774, "grad_norm": 14.524621963500977, "learning_rate": 3.834115805946792e-06, "loss": 0.0202, "step": 10230 }, { "epoch": 13.111395646606914, "grad_norm": 1.273762583732605, "learning_rate": 3.827002418551715e-06, "loss": 0.0084, "step": 10240 }, { "epoch": 13.124199743918053, "grad_norm": 10.207449913024902, "learning_rate": 3.819889031156637e-06, "loss": 0.0343, "step": 10250 }, { "epoch": 13.137003841229193, "grad_norm": 3.3465631008148193, "learning_rate": 3.8127756437615597e-06, "loss": 0.0095, "step": 10260 }, { "epoch": 13.149807938540333, "grad_norm": 0.006828089244663715, "learning_rate": 3.805662256366482e-06, "loss": 0.0155, "step": 10270 }, { "epoch": 13.162612035851472, "grad_norm": 0.49927470088005066, "learning_rate": 3.7985488689714044e-06, "loss": 0.0237, "step": 10280 }, { "epoch": 13.175416133162612, "grad_norm": 0.14710770547389984, "learning_rate": 3.7914354815763272e-06, "loss": 0.0347, "step": 10290 }, { "epoch": 13.188220230473751, "grad_norm": 0.0990980714559555, "learning_rate": 3.784322094181249e-06, "loss": 0.0134, "step": 10300 }, { "epoch": 13.201024327784891, "grad_norm": 1.8304428522242233e-05, "learning_rate": 3.777208706786172e-06, "loss": 0.0082, "step": 10310 }, { "epoch": 13.21382842509603, "grad_norm": 0.00459267245605588, "learning_rate": 3.7700953193910944e-06, "loss": 0.0036, "step": 10320 }, { "epoch": 13.22663252240717, "grad_norm": 0.054646868258714676, "learning_rate": 3.762981931996017e-06, "loss": 0.0147, "step": 10330 }, { "epoch": 13.23943661971831, "grad_norm": 4.998219013214111, "learning_rate": 3.755868544600939e-06, "loss": 0.0142, "step": 10340 }, { "epoch": 13.25224071702945, "grad_norm": 0.0036943841259926558, "learning_rate": 3.748755157205862e-06, "loss": 0.0293, "step": 10350 }, { "epoch": 13.26504481434059, "grad_norm": 6.776094436645508, "learning_rate": 3.7416417698107843e-06, "loss": 0.0273, "step": 10360 }, { "epoch": 13.277848911651729, "grad_norm": 1.106736421585083, "learning_rate": 3.734528382415707e-06, "loss": 0.0257, "step": 10370 }, { "epoch": 13.290653008962868, "grad_norm": 11.938494682312012, "learning_rate": 3.727414995020629e-06, "loss": 0.0139, "step": 10380 }, { "epoch": 13.303457106274008, "grad_norm": 3.600757918320596e-05, "learning_rate": 3.7203016076255514e-06, "loss": 0.007, "step": 10390 }, { "epoch": 13.316261203585148, "grad_norm": 0.00037831757799722254, "learning_rate": 3.713188220230474e-06, "loss": 0.0267, "step": 10400 }, { "epoch": 13.329065300896287, "grad_norm": 0.0019878472667187452, "learning_rate": 3.706074832835396e-06, "loss": 0.0197, "step": 10410 }, { "epoch": 13.341869398207427, "grad_norm": 0.0022884909994900227, "learning_rate": 3.698961445440319e-06, "loss": 0.0011, "step": 10420 }, { "epoch": 13.354673495518567, "grad_norm": 0.0027324643451720476, "learning_rate": 3.6918480580452413e-06, "loss": 0.031, "step": 10430 }, { "epoch": 13.367477592829706, "grad_norm": 19.304534912109375, "learning_rate": 3.684734670650164e-06, "loss": 0.0387, "step": 10440 }, { "epoch": 13.380281690140846, "grad_norm": 15.473051071166992, "learning_rate": 3.677621283255086e-06, "loss": 0.0168, "step": 10450 }, { "epoch": 13.393085787451984, "grad_norm": 1.490252137184143, "learning_rate": 3.670507895860009e-06, "loss": 0.0208, "step": 10460 }, { "epoch": 13.405889884763123, "grad_norm": 0.007035966496914625, "learning_rate": 3.6633945084649312e-06, "loss": 0.0136, "step": 10470 }, { "epoch": 13.418693982074263, "grad_norm": 0.003252288093790412, "learning_rate": 3.656281121069854e-06, "loss": 0.0086, "step": 10480 }, { "epoch": 13.431498079385403, "grad_norm": 0.4878418445587158, "learning_rate": 3.649167733674776e-06, "loss": 0.0392, "step": 10490 }, { "epoch": 13.444302176696542, "grad_norm": 4.22904109954834, "learning_rate": 3.6420543462796988e-06, "loss": 0.0292, "step": 10500 }, { "epoch": 13.457106274007682, "grad_norm": 0.0052803452126681805, "learning_rate": 3.634940958884621e-06, "loss": 0.0063, "step": 10510 }, { "epoch": 13.469910371318822, "grad_norm": 0.006046518217772245, "learning_rate": 3.6278275714895435e-06, "loss": 0.0339, "step": 10520 }, { "epoch": 13.482714468629961, "grad_norm": 0.0557611808180809, "learning_rate": 3.620714184094466e-06, "loss": 0.0117, "step": 10530 }, { "epoch": 13.4955185659411, "grad_norm": 2.972399950027466, "learning_rate": 3.6136007966993887e-06, "loss": 0.0181, "step": 10540 }, { "epoch": 13.50832266325224, "grad_norm": 8.456597328186035, "learning_rate": 3.606487409304311e-06, "loss": 0.0244, "step": 10550 }, { "epoch": 13.52112676056338, "grad_norm": 0.0005350671126507223, "learning_rate": 3.5993740219092334e-06, "loss": 0.0135, "step": 10560 }, { "epoch": 13.53393085787452, "grad_norm": 9.475726127624512, "learning_rate": 3.592260634514156e-06, "loss": 0.022, "step": 10570 }, { "epoch": 13.54673495518566, "grad_norm": 2.7988171577453613, "learning_rate": 3.5851472471190786e-06, "loss": 0.0151, "step": 10580 }, { "epoch": 13.559539052496799, "grad_norm": 0.006821895018219948, "learning_rate": 3.578033859724001e-06, "loss": 0.0141, "step": 10590 }, { "epoch": 13.572343149807939, "grad_norm": 8.041661262512207, "learning_rate": 3.5709204723289233e-06, "loss": 0.0332, "step": 10600 }, { "epoch": 13.585147247119078, "grad_norm": 0.00047529867151752114, "learning_rate": 3.5638070849338457e-06, "loss": 0.023, "step": 10610 }, { "epoch": 13.597951344430218, "grad_norm": 0.18414749205112457, "learning_rate": 3.5566936975387685e-06, "loss": 0.0117, "step": 10620 }, { "epoch": 13.610755441741357, "grad_norm": 5.8325395584106445, "learning_rate": 3.549580310143691e-06, "loss": 0.0273, "step": 10630 }, { "epoch": 13.623559539052497, "grad_norm": 2.0931334495544434, "learning_rate": 3.542466922748613e-06, "loss": 0.0303, "step": 10640 }, { "epoch": 13.636363636363637, "grad_norm": 0.23345398902893066, "learning_rate": 3.5353535353535356e-06, "loss": 0.0083, "step": 10650 }, { "epoch": 13.649167733674776, "grad_norm": 3.7202231884002686, "learning_rate": 3.5282401479584584e-06, "loss": 0.014, "step": 10660 }, { "epoch": 13.661971830985916, "grad_norm": 0.22626596689224243, "learning_rate": 3.5211267605633804e-06, "loss": 0.0125, "step": 10670 }, { "epoch": 13.674775928297056, "grad_norm": 16.9751033782959, "learning_rate": 3.5140133731683028e-06, "loss": 0.0347, "step": 10680 }, { "epoch": 13.687580025608195, "grad_norm": 1.9810250997543335, "learning_rate": 3.5068999857732256e-06, "loss": 0.0102, "step": 10690 }, { "epoch": 13.700384122919335, "grad_norm": 5.396705150604248, "learning_rate": 3.4997865983781484e-06, "loss": 0.0303, "step": 10700 }, { "epoch": 13.713188220230474, "grad_norm": 0.0016104003880172968, "learning_rate": 3.4926732109830703e-06, "loss": 0.0009, "step": 10710 }, { "epoch": 13.725992317541614, "grad_norm": 0.29903772473335266, "learning_rate": 3.4855598235879927e-06, "loss": 0.018, "step": 10720 }, { "epoch": 13.738796414852754, "grad_norm": 0.0008271485567092896, "learning_rate": 3.4784464361929155e-06, "loss": 0.0194, "step": 10730 }, { "epoch": 13.751600512163893, "grad_norm": 0.014146417379379272, "learning_rate": 3.4713330487978383e-06, "loss": 0.0024, "step": 10740 }, { "epoch": 13.764404609475031, "grad_norm": 1.3157802820205688, "learning_rate": 3.4642196614027602e-06, "loss": 0.0044, "step": 10750 }, { "epoch": 13.77720870678617, "grad_norm": 1.8869186639785767, "learning_rate": 3.4571062740076826e-06, "loss": 0.0254, "step": 10760 }, { "epoch": 13.79001280409731, "grad_norm": 3.0772736072540283, "learning_rate": 3.4499928866126054e-06, "loss": 0.0345, "step": 10770 }, { "epoch": 13.80281690140845, "grad_norm": 5.144647598266602, "learning_rate": 3.4428794992175273e-06, "loss": 0.0251, "step": 10780 }, { "epoch": 13.81562099871959, "grad_norm": 5.6104841232299805, "learning_rate": 3.43576611182245e-06, "loss": 0.0333, "step": 10790 }, { "epoch": 13.82842509603073, "grad_norm": 0.00017765576194506139, "learning_rate": 3.4286527244273725e-06, "loss": 0.0165, "step": 10800 }, { "epoch": 13.841229193341869, "grad_norm": 8.58414077758789, "learning_rate": 3.4215393370322953e-06, "loss": 0.0368, "step": 10810 }, { "epoch": 13.854033290653009, "grad_norm": 6.548843860626221, "learning_rate": 3.4144259496372173e-06, "loss": 0.0214, "step": 10820 }, { "epoch": 13.866837387964148, "grad_norm": 1.7748005390167236, "learning_rate": 3.40731256224214e-06, "loss": 0.0105, "step": 10830 }, { "epoch": 13.879641485275288, "grad_norm": 0.02361682802438736, "learning_rate": 3.4001991748470624e-06, "loss": 0.0039, "step": 10840 }, { "epoch": 13.892445582586427, "grad_norm": 0.0009472609381191432, "learning_rate": 3.3930857874519852e-06, "loss": 0.0244, "step": 10850 }, { "epoch": 13.905249679897567, "grad_norm": 8.113651275634766, "learning_rate": 3.385972400056907e-06, "loss": 0.0288, "step": 10860 }, { "epoch": 13.918053777208707, "grad_norm": 0.03320278227329254, "learning_rate": 3.37885901266183e-06, "loss": 0.0196, "step": 10870 }, { "epoch": 13.930857874519846, "grad_norm": 0.04235280677676201, "learning_rate": 3.3717456252667523e-06, "loss": 0.0134, "step": 10880 }, { "epoch": 13.943661971830986, "grad_norm": 4.171895980834961, "learning_rate": 3.364632237871675e-06, "loss": 0.0178, "step": 10890 }, { "epoch": 13.956466069142126, "grad_norm": 0.032928042113780975, "learning_rate": 3.357518850476597e-06, "loss": 0.0142, "step": 10900 }, { "epoch": 13.969270166453265, "grad_norm": 0.0006852024816907942, "learning_rate": 3.35040546308152e-06, "loss": 0.0055, "step": 10910 }, { "epoch": 13.982074263764405, "grad_norm": 0.0023214409593492746, "learning_rate": 3.3432920756864423e-06, "loss": 0.0116, "step": 10920 }, { "epoch": 13.994878361075545, "grad_norm": 0.38349929451942444, "learning_rate": 3.336178688291364e-06, "loss": 0.0235, "step": 10930 }, { "epoch": 14.007682458386684, "grad_norm": 2.070441961288452, "learning_rate": 3.329065300896287e-06, "loss": 0.0086, "step": 10940 }, { "epoch": 14.020486555697824, "grad_norm": 0.12442722916603088, "learning_rate": 3.32195191350121e-06, "loss": 0.0072, "step": 10950 }, { "epoch": 14.033290653008963, "grad_norm": 6.558221340179443, "learning_rate": 3.314838526106132e-06, "loss": 0.0252, "step": 10960 }, { "epoch": 14.046094750320103, "grad_norm": 5.9624271671054885e-05, "learning_rate": 3.307725138711054e-06, "loss": 0.0193, "step": 10970 }, { "epoch": 14.058898847631243, "grad_norm": 0.0025486797094345093, "learning_rate": 3.300611751315977e-06, "loss": 0.0197, "step": 10980 }, { "epoch": 14.071702944942382, "grad_norm": 4.4869384765625, "learning_rate": 3.2934983639208993e-06, "loss": 0.0007, "step": 10990 }, { "epoch": 14.084507042253522, "grad_norm": 7.89267651271075e-05, "learning_rate": 3.286384976525822e-06, "loss": 0.0151, "step": 11000 }, { "epoch": 14.097311139564662, "grad_norm": 13.062664031982422, "learning_rate": 3.279271589130744e-06, "loss": 0.0189, "step": 11010 }, { "epoch": 14.1101152368758, "grad_norm": 0.5844500064849854, "learning_rate": 3.272158201735667e-06, "loss": 0.0028, "step": 11020 }, { "epoch": 14.122919334186939, "grad_norm": 16.713825225830078, "learning_rate": 3.265044814340589e-06, "loss": 0.044, "step": 11030 }, { "epoch": 14.135723431498079, "grad_norm": 5.629806995391846, "learning_rate": 3.2579314269455116e-06, "loss": 0.0359, "step": 11040 }, { "epoch": 14.148527528809218, "grad_norm": 5.0279154777526855, "learning_rate": 3.250818039550434e-06, "loss": 0.0398, "step": 11050 }, { "epoch": 14.161331626120358, "grad_norm": 14.42271900177002, "learning_rate": 3.2437046521553568e-06, "loss": 0.0339, "step": 11060 }, { "epoch": 14.174135723431498, "grad_norm": 2.7477522962726653e-05, "learning_rate": 3.236591264760279e-06, "loss": 0.0414, "step": 11070 }, { "epoch": 14.186939820742637, "grad_norm": 0.005621842574328184, "learning_rate": 3.2294778773652015e-06, "loss": 0.0099, "step": 11080 }, { "epoch": 14.199743918053777, "grad_norm": 0.00278816232457757, "learning_rate": 3.222364489970124e-06, "loss": 0.0144, "step": 11090 }, { "epoch": 14.212548015364916, "grad_norm": 0.005414939485490322, "learning_rate": 3.2152511025750467e-06, "loss": 0.008, "step": 11100 }, { "epoch": 14.225352112676056, "grad_norm": 0.21644172072410583, "learning_rate": 3.208137715179969e-06, "loss": 0.0127, "step": 11110 }, { "epoch": 14.238156209987196, "grad_norm": 0.0003669892903417349, "learning_rate": 3.2010243277848914e-06, "loss": 0.0037, "step": 11120 }, { "epoch": 14.250960307298335, "grad_norm": 10.803598403930664, "learning_rate": 3.1939109403898138e-06, "loss": 0.0176, "step": 11130 }, { "epoch": 14.263764404609475, "grad_norm": 0.00011865006672451273, "learning_rate": 3.1867975529947366e-06, "loss": 0.0136, "step": 11140 }, { "epoch": 14.276568501920615, "grad_norm": 7.826897144317627, "learning_rate": 3.179684165599659e-06, "loss": 0.0189, "step": 11150 }, { "epoch": 14.289372599231754, "grad_norm": 0.02872617356479168, "learning_rate": 3.1725707782045813e-06, "loss": 0.0281, "step": 11160 }, { "epoch": 14.302176696542894, "grad_norm": 6.168277740478516, "learning_rate": 3.1654573908095037e-06, "loss": 0.0091, "step": 11170 }, { "epoch": 14.314980793854033, "grad_norm": 0.00011484704009490088, "learning_rate": 3.1583440034144265e-06, "loss": 0.0185, "step": 11180 }, { "epoch": 14.327784891165173, "grad_norm": 0.02605835348367691, "learning_rate": 3.1512306160193484e-06, "loss": 0.029, "step": 11190 }, { "epoch": 14.340588988476313, "grad_norm": 5.659068584442139, "learning_rate": 3.1441172286242712e-06, "loss": 0.024, "step": 11200 }, { "epoch": 14.353393085787452, "grad_norm": 4.635328769683838, "learning_rate": 3.1370038412291936e-06, "loss": 0.0103, "step": 11210 }, { "epoch": 14.366197183098592, "grad_norm": 1.1552577018737793, "learning_rate": 3.1298904538341164e-06, "loss": 0.0354, "step": 11220 }, { "epoch": 14.379001280409732, "grad_norm": 7.550588130950928, "learning_rate": 3.1227770664390384e-06, "loss": 0.0192, "step": 11230 }, { "epoch": 14.391805377720871, "grad_norm": 0.00020585409947670996, "learning_rate": 3.1156636790439607e-06, "loss": 0.0096, "step": 11240 }, { "epoch": 14.40460947503201, "grad_norm": 9.821633338928223, "learning_rate": 3.1085502916488835e-06, "loss": 0.0197, "step": 11250 }, { "epoch": 14.41741357234315, "grad_norm": 1.478725880588172e-05, "learning_rate": 3.1014369042538063e-06, "loss": 0.0189, "step": 11260 }, { "epoch": 14.43021766965429, "grad_norm": 5.791157245635986, "learning_rate": 3.0943235168587283e-06, "loss": 0.0212, "step": 11270 }, { "epoch": 14.44302176696543, "grad_norm": 0.005984743125736713, "learning_rate": 3.0872101294636507e-06, "loss": 0.0055, "step": 11280 }, { "epoch": 14.45582586427657, "grad_norm": 8.674723625183105, "learning_rate": 3.0800967420685735e-06, "loss": 0.0464, "step": 11290 }, { "epoch": 14.468629961587709, "grad_norm": 0.013488815166056156, "learning_rate": 3.0729833546734954e-06, "loss": 0.0029, "step": 11300 }, { "epoch": 14.481434058898847, "grad_norm": 0.0004084960964974016, "learning_rate": 3.065869967278418e-06, "loss": 0.0073, "step": 11310 }, { "epoch": 14.494238156209986, "grad_norm": 8.93881893157959, "learning_rate": 3.0587565798833406e-06, "loss": 0.0484, "step": 11320 }, { "epoch": 14.507042253521126, "grad_norm": 0.0002386833803029731, "learning_rate": 3.0516431924882634e-06, "loss": 0.0016, "step": 11330 }, { "epoch": 14.519846350832266, "grad_norm": 4.354171276092529, "learning_rate": 3.0445298050931853e-06, "loss": 0.0218, "step": 11340 }, { "epoch": 14.532650448143405, "grad_norm": 3.054635763168335, "learning_rate": 3.037416417698108e-06, "loss": 0.0108, "step": 11350 }, { "epoch": 14.545454545454545, "grad_norm": 2.393022060394287, "learning_rate": 3.0303030303030305e-06, "loss": 0.0257, "step": 11360 }, { "epoch": 14.558258642765685, "grad_norm": 1.033687949180603, "learning_rate": 3.0231896429079533e-06, "loss": 0.0048, "step": 11370 }, { "epoch": 14.571062740076824, "grad_norm": 3.4994325637817383, "learning_rate": 3.0160762555128752e-06, "loss": 0.0019, "step": 11380 }, { "epoch": 14.583866837387964, "grad_norm": 15.026253700256348, "learning_rate": 3.008962868117798e-06, "loss": 0.0425, "step": 11390 }, { "epoch": 14.596670934699103, "grad_norm": 0.008529364131391048, "learning_rate": 3.0018494807227204e-06, "loss": 0.0194, "step": 11400 }, { "epoch": 14.609475032010243, "grad_norm": 0.17839021980762482, "learning_rate": 2.994736093327643e-06, "loss": 0.0212, "step": 11410 }, { "epoch": 14.622279129321383, "grad_norm": 0.05387525260448456, "learning_rate": 2.987622705932565e-06, "loss": 0.0361, "step": 11420 }, { "epoch": 14.635083226632522, "grad_norm": 0.0007083885720930994, "learning_rate": 2.980509318537488e-06, "loss": 0.0449, "step": 11430 }, { "epoch": 14.647887323943662, "grad_norm": 0.004359661601483822, "learning_rate": 2.9733959311424103e-06, "loss": 0.0238, "step": 11440 }, { "epoch": 14.660691421254802, "grad_norm": 0.8683837056159973, "learning_rate": 2.9662825437473327e-06, "loss": 0.0097, "step": 11450 }, { "epoch": 14.673495518565941, "grad_norm": 0.0002977359981741756, "learning_rate": 2.959169156352255e-06, "loss": 0.0532, "step": 11460 }, { "epoch": 14.68629961587708, "grad_norm": 8.213149070739746, "learning_rate": 2.952055768957178e-06, "loss": 0.0301, "step": 11470 }, { "epoch": 14.69910371318822, "grad_norm": 0.10097076743841171, "learning_rate": 2.9449423815621002e-06, "loss": 0.0167, "step": 11480 }, { "epoch": 14.71190781049936, "grad_norm": 1.4323228597640991, "learning_rate": 2.937828994167022e-06, "loss": 0.0213, "step": 11490 }, { "epoch": 14.7247119078105, "grad_norm": 0.3369174003601074, "learning_rate": 2.930715606771945e-06, "loss": 0.0091, "step": 11500 }, { "epoch": 14.73751600512164, "grad_norm": 0.033116430044174194, "learning_rate": 2.9236022193768678e-06, "loss": 0.0341, "step": 11510 }, { "epoch": 14.750320102432779, "grad_norm": 0.053776051849126816, "learning_rate": 2.91648883198179e-06, "loss": 0.006, "step": 11520 }, { "epoch": 14.763124199743919, "grad_norm": 0.0004347638168837875, "learning_rate": 2.909375444586712e-06, "loss": 0.0095, "step": 11530 }, { "epoch": 14.775928297055058, "grad_norm": 0.004032296594232321, "learning_rate": 2.902262057191635e-06, "loss": 0.0076, "step": 11540 }, { "epoch": 14.788732394366198, "grad_norm": 14.899596214294434, "learning_rate": 2.8951486697965577e-06, "loss": 0.0115, "step": 11550 }, { "epoch": 14.801536491677338, "grad_norm": 0.0006039092550054193, "learning_rate": 2.8880352824014796e-06, "loss": 0.0011, "step": 11560 }, { "epoch": 14.814340588988475, "grad_norm": 5.994747638702393, "learning_rate": 2.880921895006402e-06, "loss": 0.0054, "step": 11570 }, { "epoch": 14.827144686299615, "grad_norm": 0.2780959904193878, "learning_rate": 2.873808507611325e-06, "loss": 0.0061, "step": 11580 }, { "epoch": 14.839948783610755, "grad_norm": 4.8738226890563965, "learning_rate": 2.8666951202162476e-06, "loss": 0.0272, "step": 11590 }, { "epoch": 14.852752880921894, "grad_norm": 0.014950653538107872, "learning_rate": 2.8595817328211696e-06, "loss": 0.0066, "step": 11600 }, { "epoch": 14.865556978233034, "grad_norm": 8.080557823181152, "learning_rate": 2.852468345426092e-06, "loss": 0.0162, "step": 11610 }, { "epoch": 14.878361075544174, "grad_norm": 0.0905604362487793, "learning_rate": 2.8453549580310147e-06, "loss": 0.0005, "step": 11620 }, { "epoch": 14.891165172855313, "grad_norm": 1.771815538406372, "learning_rate": 2.838241570635937e-06, "loss": 0.018, "step": 11630 }, { "epoch": 14.903969270166453, "grad_norm": 19.153865814208984, "learning_rate": 2.8311281832408595e-06, "loss": 0.0282, "step": 11640 }, { "epoch": 14.916773367477592, "grad_norm": 11.782116889953613, "learning_rate": 2.824014795845782e-06, "loss": 0.0382, "step": 11650 }, { "epoch": 14.929577464788732, "grad_norm": 0.04391075670719147, "learning_rate": 2.8169014084507046e-06, "loss": 0.0022, "step": 11660 }, { "epoch": 14.942381562099872, "grad_norm": 2.7921619415283203, "learning_rate": 2.8097880210556266e-06, "loss": 0.0243, "step": 11670 }, { "epoch": 14.955185659411011, "grad_norm": 3.0086295737419277e-05, "learning_rate": 2.8026746336605494e-06, "loss": 0.0312, "step": 11680 }, { "epoch": 14.967989756722151, "grad_norm": 11.096000671386719, "learning_rate": 2.7955612462654718e-06, "loss": 0.027, "step": 11690 }, { "epoch": 14.98079385403329, "grad_norm": 0.002772057428956032, "learning_rate": 2.7884478588703946e-06, "loss": 0.0163, "step": 11700 }, { "epoch": 14.99359795134443, "grad_norm": 0.0588461272418499, "learning_rate": 2.7813344714753165e-06, "loss": 0.0085, "step": 11710 }, { "epoch": 15.00640204865557, "grad_norm": 0.00016617824439890683, "learning_rate": 2.7742210840802393e-06, "loss": 0.0058, "step": 11720 }, { "epoch": 15.01920614596671, "grad_norm": 0.019842274487018585, "learning_rate": 2.7671076966851617e-06, "loss": 0.0281, "step": 11730 }, { "epoch": 15.032010243277849, "grad_norm": 0.00029976878431625664, "learning_rate": 2.7599943092900845e-06, "loss": 0.0167, "step": 11740 }, { "epoch": 15.044814340588989, "grad_norm": 5.000084400177002, "learning_rate": 2.7528809218950064e-06, "loss": 0.0126, "step": 11750 }, { "epoch": 15.057618437900128, "grad_norm": 5.378047466278076, "learning_rate": 2.7457675344999292e-06, "loss": 0.0221, "step": 11760 }, { "epoch": 15.070422535211268, "grad_norm": 0.34347859025001526, "learning_rate": 2.7386541471048516e-06, "loss": 0.0081, "step": 11770 }, { "epoch": 15.083226632522408, "grad_norm": 0.003625217592343688, "learning_rate": 2.7315407597097744e-06, "loss": 0.0169, "step": 11780 }, { "epoch": 15.096030729833547, "grad_norm": 2.1973159313201904, "learning_rate": 2.7244273723146963e-06, "loss": 0.0092, "step": 11790 }, { "epoch": 15.108834827144687, "grad_norm": 0.16450557112693787, "learning_rate": 2.717313984919619e-06, "loss": 0.0204, "step": 11800 }, { "epoch": 15.121638924455826, "grad_norm": 0.2625795900821686, "learning_rate": 2.7102005975245415e-06, "loss": 0.0344, "step": 11810 }, { "epoch": 15.134443021766966, "grad_norm": 0.0039482396095991135, "learning_rate": 2.7030872101294635e-06, "loss": 0.0101, "step": 11820 }, { "epoch": 15.147247119078106, "grad_norm": 0.00013273139484226704, "learning_rate": 2.6959738227343863e-06, "loss": 0.0308, "step": 11830 }, { "epoch": 15.160051216389245, "grad_norm": 0.012734783813357353, "learning_rate": 2.688860435339309e-06, "loss": 0.008, "step": 11840 }, { "epoch": 15.172855313700385, "grad_norm": 10.068683624267578, "learning_rate": 2.6817470479442314e-06, "loss": 0.0297, "step": 11850 }, { "epoch": 15.185659411011523, "grad_norm": 0.001013003522530198, "learning_rate": 2.6746336605491534e-06, "loss": 0.0127, "step": 11860 }, { "epoch": 15.198463508322662, "grad_norm": 1.4995566606521606, "learning_rate": 2.667520273154076e-06, "loss": 0.0159, "step": 11870 }, { "epoch": 15.211267605633802, "grad_norm": 0.4025968313217163, "learning_rate": 2.6604068857589986e-06, "loss": 0.0055, "step": 11880 }, { "epoch": 15.224071702944942, "grad_norm": 1.3007484674453735, "learning_rate": 2.6532934983639213e-06, "loss": 0.0197, "step": 11890 }, { "epoch": 15.236875800256081, "grad_norm": 0.003301259595900774, "learning_rate": 2.6461801109688433e-06, "loss": 0.0212, "step": 11900 }, { "epoch": 15.249679897567221, "grad_norm": 0.0012052332749590278, "learning_rate": 2.639066723573766e-06, "loss": 0.0074, "step": 11910 }, { "epoch": 15.26248399487836, "grad_norm": 1.908300757408142, "learning_rate": 2.6319533361786885e-06, "loss": 0.0264, "step": 11920 }, { "epoch": 15.2752880921895, "grad_norm": 0.000183357551577501, "learning_rate": 2.624839948783611e-06, "loss": 0.0072, "step": 11930 }, { "epoch": 15.28809218950064, "grad_norm": 0.004172053188085556, "learning_rate": 2.6177265613885332e-06, "loss": 0.0045, "step": 11940 }, { "epoch": 15.30089628681178, "grad_norm": 0.2952759265899658, "learning_rate": 2.610613173993456e-06, "loss": 0.0178, "step": 11950 }, { "epoch": 15.31370038412292, "grad_norm": 0.004701425787061453, "learning_rate": 2.6034997865983784e-06, "loss": 0.0053, "step": 11960 }, { "epoch": 15.326504481434059, "grad_norm": 9.080245971679688, "learning_rate": 2.5963863992033008e-06, "loss": 0.0409, "step": 11970 }, { "epoch": 15.339308578745198, "grad_norm": 0.08763319253921509, "learning_rate": 2.589273011808223e-06, "loss": 0.0158, "step": 11980 }, { "epoch": 15.352112676056338, "grad_norm": 7.685142993927002, "learning_rate": 2.582159624413146e-06, "loss": 0.0224, "step": 11990 }, { "epoch": 15.364916773367478, "grad_norm": 0.0004092796880286187, "learning_rate": 2.5750462370180683e-06, "loss": 0.0178, "step": 12000 }, { "epoch": 15.377720870678617, "grad_norm": 0.3881383240222931, "learning_rate": 2.5679328496229907e-06, "loss": 0.0139, "step": 12010 }, { "epoch": 15.390524967989757, "grad_norm": 6.461103439331055, "learning_rate": 2.560819462227913e-06, "loss": 0.0232, "step": 12020 }, { "epoch": 15.403329065300897, "grad_norm": 0.0015170153928920627, "learning_rate": 2.553706074832836e-06, "loss": 0.0196, "step": 12030 }, { "epoch": 15.416133162612036, "grad_norm": 0.011582997627556324, "learning_rate": 2.5465926874377582e-06, "loss": 0.023, "step": 12040 }, { "epoch": 15.428937259923176, "grad_norm": 0.40539681911468506, "learning_rate": 2.5394793000426806e-06, "loss": 0.0125, "step": 12050 }, { "epoch": 15.441741357234315, "grad_norm": 2.6545543670654297, "learning_rate": 2.532365912647603e-06, "loss": 0.0128, "step": 12060 }, { "epoch": 15.454545454545455, "grad_norm": 0.0018354145577177405, "learning_rate": 2.5252525252525258e-06, "loss": 0.0321, "step": 12070 }, { "epoch": 15.467349551856595, "grad_norm": 0.0012273455504328012, "learning_rate": 2.5181391378574477e-06, "loss": 0.0095, "step": 12080 }, { "epoch": 15.480153649167734, "grad_norm": 0.001059592585079372, "learning_rate": 2.51102575046237e-06, "loss": 0.0349, "step": 12090 }, { "epoch": 15.492957746478874, "grad_norm": 0.5970710515975952, "learning_rate": 2.503912363067293e-06, "loss": 0.0058, "step": 12100 }, { "epoch": 15.505761843790014, "grad_norm": 0.012427791021764278, "learning_rate": 2.4967989756722153e-06, "loss": 0.0157, "step": 12110 }, { "epoch": 15.518565941101153, "grad_norm": 6.88558939145878e-05, "learning_rate": 2.489685588277138e-06, "loss": 0.0183, "step": 12120 }, { "epoch": 15.531370038412291, "grad_norm": 0.3271321654319763, "learning_rate": 2.48257220088206e-06, "loss": 0.0135, "step": 12130 }, { "epoch": 15.54417413572343, "grad_norm": 1.5439924001693726, "learning_rate": 2.475458813486983e-06, "loss": 0.0054, "step": 12140 }, { "epoch": 15.55697823303457, "grad_norm": 0.0002505710581317544, "learning_rate": 2.468345426091905e-06, "loss": 0.005, "step": 12150 }, { "epoch": 15.56978233034571, "grad_norm": 0.22347617149353027, "learning_rate": 2.4612320386968275e-06, "loss": 0.0567, "step": 12160 }, { "epoch": 15.58258642765685, "grad_norm": 2.8955597372259945e-05, "learning_rate": 2.45411865130175e-06, "loss": 0.0047, "step": 12170 }, { "epoch": 15.59539052496799, "grad_norm": 0.005480688996613026, "learning_rate": 2.4470052639066723e-06, "loss": 0.0089, "step": 12180 }, { "epoch": 15.608194622279129, "grad_norm": 0.9169988036155701, "learning_rate": 2.439891876511595e-06, "loss": 0.0071, "step": 12190 }, { "epoch": 15.620998719590268, "grad_norm": 0.07415187358856201, "learning_rate": 2.4327784891165175e-06, "loss": 0.0255, "step": 12200 }, { "epoch": 15.633802816901408, "grad_norm": 10.148873329162598, "learning_rate": 2.42566510172144e-06, "loss": 0.0095, "step": 12210 }, { "epoch": 15.646606914212548, "grad_norm": 0.7196668982505798, "learning_rate": 2.418551714326362e-06, "loss": 0.0018, "step": 12220 }, { "epoch": 15.659411011523687, "grad_norm": 0.03670328110456467, "learning_rate": 2.411438326931285e-06, "loss": 0.018, "step": 12230 }, { "epoch": 15.672215108834827, "grad_norm": 0.0001274590176763013, "learning_rate": 2.4043249395362074e-06, "loss": 0.0384, "step": 12240 }, { "epoch": 15.685019206145967, "grad_norm": 0.2096865028142929, "learning_rate": 2.3972115521411297e-06, "loss": 0.0271, "step": 12250 }, { "epoch": 15.697823303457106, "grad_norm": 3.228276545996778e-05, "learning_rate": 2.390098164746052e-06, "loss": 0.0083, "step": 12260 }, { "epoch": 15.710627400768246, "grad_norm": 0.013042661361396313, "learning_rate": 2.382984777350975e-06, "loss": 0.0071, "step": 12270 }, { "epoch": 15.723431498079385, "grad_norm": 0.0041104513220489025, "learning_rate": 2.3758713899558973e-06, "loss": 0.0007, "step": 12280 }, { "epoch": 15.736235595390525, "grad_norm": 9.80040168762207, "learning_rate": 2.3687580025608197e-06, "loss": 0.0063, "step": 12290 }, { "epoch": 15.749039692701665, "grad_norm": 0.7197008728981018, "learning_rate": 2.361644615165742e-06, "loss": 0.0201, "step": 12300 }, { "epoch": 15.761843790012804, "grad_norm": 0.14606595039367676, "learning_rate": 2.3545312277706644e-06, "loss": 0.0043, "step": 12310 }, { "epoch": 15.774647887323944, "grad_norm": 0.21366263926029205, "learning_rate": 2.347417840375587e-06, "loss": 0.0146, "step": 12320 }, { "epoch": 15.787451984635084, "grad_norm": 0.005361586343497038, "learning_rate": 2.3403044529805096e-06, "loss": 0.0403, "step": 12330 }, { "epoch": 15.800256081946223, "grad_norm": 9.570905240252614e-05, "learning_rate": 2.333191065585432e-06, "loss": 0.012, "step": 12340 }, { "epoch": 15.813060179257363, "grad_norm": 0.50851970911026, "learning_rate": 2.3260776781903543e-06, "loss": 0.0151, "step": 12350 }, { "epoch": 15.825864276568502, "grad_norm": 0.013562064617872238, "learning_rate": 2.318964290795277e-06, "loss": 0.0064, "step": 12360 }, { "epoch": 15.838668373879642, "grad_norm": 0.0003261865640524775, "learning_rate": 2.3118509034001995e-06, "loss": 0.0021, "step": 12370 }, { "epoch": 15.851472471190782, "grad_norm": 4.786178112030029, "learning_rate": 2.304737516005122e-06, "loss": 0.0072, "step": 12380 }, { "epoch": 15.864276568501921, "grad_norm": 0.011111567728221416, "learning_rate": 2.2976241286100442e-06, "loss": 0.0166, "step": 12390 }, { "epoch": 15.877080665813061, "grad_norm": 1.9213408450013958e-05, "learning_rate": 2.290510741214967e-06, "loss": 0.0207, "step": 12400 }, { "epoch": 15.8898847631242, "grad_norm": 0.09700735658407211, "learning_rate": 2.283397353819889e-06, "loss": 0.0115, "step": 12410 }, { "epoch": 15.90268886043534, "grad_norm": 1.992964506149292, "learning_rate": 2.2762839664248118e-06, "loss": 0.0285, "step": 12420 }, { "epoch": 15.915492957746478, "grad_norm": 0.3744063377380371, "learning_rate": 2.269170579029734e-06, "loss": 0.0117, "step": 12430 }, { "epoch": 15.928297055057618, "grad_norm": 0.017899269238114357, "learning_rate": 2.2620571916346565e-06, "loss": 0.0202, "step": 12440 }, { "epoch": 15.941101152368757, "grad_norm": 0.02544565685093403, "learning_rate": 2.254943804239579e-06, "loss": 0.0022, "step": 12450 }, { "epoch": 15.953905249679897, "grad_norm": 0.7286255955696106, "learning_rate": 2.2478304168445013e-06, "loss": 0.0129, "step": 12460 }, { "epoch": 15.966709346991037, "grad_norm": 0.00012173001596238464, "learning_rate": 2.240717029449424e-06, "loss": 0.0198, "step": 12470 }, { "epoch": 15.979513444302176, "grad_norm": 3.1849424839019775, "learning_rate": 2.2336036420543464e-06, "loss": 0.0193, "step": 12480 }, { "epoch": 15.992317541613316, "grad_norm": 0.35744091868400574, "learning_rate": 2.226490254659269e-06, "loss": 0.0179, "step": 12490 }, { "epoch": 16.005121638924457, "grad_norm": 1.168373703956604, "learning_rate": 2.219376867264191e-06, "loss": 0.0052, "step": 12500 }, { "epoch": 16.017925736235597, "grad_norm": 3.629526138305664, "learning_rate": 2.212263479869114e-06, "loss": 0.0021, "step": 12510 }, { "epoch": 16.030729833546737, "grad_norm": 2.5898404121398926, "learning_rate": 2.2051500924740364e-06, "loss": 0.0044, "step": 12520 }, { "epoch": 16.043533930857876, "grad_norm": 0.10092195868492126, "learning_rate": 2.1980367050789587e-06, "loss": 0.0057, "step": 12530 }, { "epoch": 16.056338028169016, "grad_norm": 0.0005303798825480044, "learning_rate": 2.190923317683881e-06, "loss": 0.036, "step": 12540 }, { "epoch": 16.069142125480155, "grad_norm": 12.291526794433594, "learning_rate": 2.183809930288804e-06, "loss": 0.0088, "step": 12550 }, { "epoch": 16.08194622279129, "grad_norm": 0.0005587077466771007, "learning_rate": 2.1766965428937263e-06, "loss": 0.0179, "step": 12560 }, { "epoch": 16.09475032010243, "grad_norm": 5.582305908203125, "learning_rate": 2.1695831554986487e-06, "loss": 0.0132, "step": 12570 }, { "epoch": 16.10755441741357, "grad_norm": 0.003146237228065729, "learning_rate": 2.162469768103571e-06, "loss": 0.0165, "step": 12580 }, { "epoch": 16.12035851472471, "grad_norm": 0.002655381802469492, "learning_rate": 2.1553563807084934e-06, "loss": 0.0068, "step": 12590 }, { "epoch": 16.13316261203585, "grad_norm": 0.7208670973777771, "learning_rate": 2.148242993313416e-06, "loss": 0.0406, "step": 12600 }, { "epoch": 16.14596670934699, "grad_norm": 0.02880147099494934, "learning_rate": 2.1411296059183386e-06, "loss": 0.0238, "step": 12610 }, { "epoch": 16.15877080665813, "grad_norm": 5.680818557739258, "learning_rate": 2.134016218523261e-06, "loss": 0.0146, "step": 12620 }, { "epoch": 16.17157490396927, "grad_norm": 0.019193559885025024, "learning_rate": 2.1269028311281833e-06, "loss": 0.0068, "step": 12630 }, { "epoch": 16.18437900128041, "grad_norm": 9.778482437133789, "learning_rate": 2.119789443733106e-06, "loss": 0.0214, "step": 12640 }, { "epoch": 16.197183098591548, "grad_norm": 0.02019951306283474, "learning_rate": 2.1126760563380285e-06, "loss": 0.0212, "step": 12650 }, { "epoch": 16.209987195902688, "grad_norm": 4.547806739807129, "learning_rate": 2.105562668942951e-06, "loss": 0.0192, "step": 12660 }, { "epoch": 16.222791293213827, "grad_norm": 0.0004286708135623485, "learning_rate": 2.0984492815478732e-06, "loss": 0.012, "step": 12670 }, { "epoch": 16.235595390524967, "grad_norm": 0.0002181783929700032, "learning_rate": 2.091335894152796e-06, "loss": 0.0138, "step": 12680 }, { "epoch": 16.248399487836107, "grad_norm": 0.35543304681777954, "learning_rate": 2.0842225067577184e-06, "loss": 0.016, "step": 12690 }, { "epoch": 16.261203585147246, "grad_norm": 6.514838218688965, "learning_rate": 2.0771091193626404e-06, "loss": 0.0068, "step": 12700 }, { "epoch": 16.274007682458386, "grad_norm": 6.154721736907959, "learning_rate": 2.069995731967563e-06, "loss": 0.013, "step": 12710 }, { "epoch": 16.286811779769526, "grad_norm": 4.94442892074585, "learning_rate": 2.0628823445724855e-06, "loss": 0.0048, "step": 12720 }, { "epoch": 16.299615877080665, "grad_norm": 0.5289265513420105, "learning_rate": 2.055768957177408e-06, "loss": 0.0021, "step": 12730 }, { "epoch": 16.312419974391805, "grad_norm": 1.8853663206100464, "learning_rate": 2.0486555697823303e-06, "loss": 0.0181, "step": 12740 }, { "epoch": 16.325224071702944, "grad_norm": 0.0020349062979221344, "learning_rate": 2.041542182387253e-06, "loss": 0.0172, "step": 12750 }, { "epoch": 16.338028169014084, "grad_norm": 7.857889652252197, "learning_rate": 2.0344287949921754e-06, "loss": 0.0079, "step": 12760 }, { "epoch": 16.350832266325224, "grad_norm": 0.7145957946777344, "learning_rate": 2.027315407597098e-06, "loss": 0.0074, "step": 12770 }, { "epoch": 16.363636363636363, "grad_norm": 7.4408781074453145e-06, "learning_rate": 2.02020202020202e-06, "loss": 0.0035, "step": 12780 }, { "epoch": 16.376440460947503, "grad_norm": 0.4833701252937317, "learning_rate": 2.013088632806943e-06, "loss": 0.0306, "step": 12790 }, { "epoch": 16.389244558258643, "grad_norm": 3.594839290599339e-05, "learning_rate": 2.0059752454118654e-06, "loss": 0.0032, "step": 12800 }, { "epoch": 16.402048655569782, "grad_norm": 0.13474607467651367, "learning_rate": 1.9988618580167877e-06, "loss": 0.0007, "step": 12810 }, { "epoch": 16.414852752880922, "grad_norm": 5.6487537221983075e-05, "learning_rate": 1.99174847062171e-06, "loss": 0.0353, "step": 12820 }, { "epoch": 16.42765685019206, "grad_norm": 0.0008355234167538583, "learning_rate": 1.9846350832266325e-06, "loss": 0.007, "step": 12830 }, { "epoch": 16.4404609475032, "grad_norm": 0.00018407643074169755, "learning_rate": 1.9775216958315553e-06, "loss": 0.0069, "step": 12840 }, { "epoch": 16.45326504481434, "grad_norm": 4.728601455688477, "learning_rate": 1.9704083084364776e-06, "loss": 0.0213, "step": 12850 }, { "epoch": 16.46606914212548, "grad_norm": 5.737601280212402, "learning_rate": 1.9632949210414e-06, "loss": 0.0324, "step": 12860 }, { "epoch": 16.47887323943662, "grad_norm": 5.883288860321045, "learning_rate": 1.9561815336463224e-06, "loss": 0.0195, "step": 12870 }, { "epoch": 16.49167733674776, "grad_norm": 7.718067169189453, "learning_rate": 1.949068146251245e-06, "loss": 0.0286, "step": 12880 }, { "epoch": 16.5044814340589, "grad_norm": 0.00023551759659312665, "learning_rate": 1.9419547588561676e-06, "loss": 0.0332, "step": 12890 }, { "epoch": 16.51728553137004, "grad_norm": 0.13080836832523346, "learning_rate": 1.93484137146109e-06, "loss": 0.0111, "step": 12900 }, { "epoch": 16.53008962868118, "grad_norm": 3.4559757709503174, "learning_rate": 1.9277279840660123e-06, "loss": 0.0147, "step": 12910 }, { "epoch": 16.542893725992318, "grad_norm": 5.610240459442139, "learning_rate": 1.920614596670935e-06, "loss": 0.0179, "step": 12920 }, { "epoch": 16.555697823303458, "grad_norm": 0.004110823851078749, "learning_rate": 1.9135012092758575e-06, "loss": 0.0215, "step": 12930 }, { "epoch": 16.568501920614597, "grad_norm": 0.03885205462574959, "learning_rate": 1.9063878218807798e-06, "loss": 0.019, "step": 12940 }, { "epoch": 16.581306017925737, "grad_norm": 0.004340150393545628, "learning_rate": 1.8992744344857022e-06, "loss": 0.0116, "step": 12950 }, { "epoch": 16.594110115236877, "grad_norm": 20.883872985839844, "learning_rate": 1.8921610470906246e-06, "loss": 0.0368, "step": 12960 }, { "epoch": 16.606914212548016, "grad_norm": 0.1189589723944664, "learning_rate": 1.8850476596955472e-06, "loss": 0.0133, "step": 12970 }, { "epoch": 16.619718309859156, "grad_norm": 5.816783905029297, "learning_rate": 1.8779342723004696e-06, "loss": 0.0083, "step": 12980 }, { "epoch": 16.632522407170296, "grad_norm": 2.225306749343872, "learning_rate": 1.8708208849053921e-06, "loss": 0.0134, "step": 12990 }, { "epoch": 16.645326504481435, "grad_norm": 3.729928266693605e-06, "learning_rate": 1.8637074975103145e-06, "loss": 0.0086, "step": 13000 }, { "epoch": 16.658130601792575, "grad_norm": 0.26913100481033325, "learning_rate": 1.856594110115237e-06, "loss": 0.0252, "step": 13010 }, { "epoch": 16.670934699103714, "grad_norm": 1.1184187314938754e-05, "learning_rate": 1.8494807227201595e-06, "loss": 0.0303, "step": 13020 }, { "epoch": 16.683738796414854, "grad_norm": 0.015479632653295994, "learning_rate": 1.842367335325082e-06, "loss": 0.0067, "step": 13030 }, { "epoch": 16.696542893725994, "grad_norm": 0.6508268117904663, "learning_rate": 1.8352539479300044e-06, "loss": 0.0163, "step": 13040 }, { "epoch": 16.709346991037133, "grad_norm": 0.00032824071240611374, "learning_rate": 1.828140560534927e-06, "loss": 0.0152, "step": 13050 }, { "epoch": 16.722151088348273, "grad_norm": 5.615657806396484, "learning_rate": 1.8210271731398494e-06, "loss": 0.0248, "step": 13060 }, { "epoch": 16.734955185659413, "grad_norm": 0.0006533341947942972, "learning_rate": 1.8139137857447718e-06, "loss": 0.012, "step": 13070 }, { "epoch": 16.747759282970552, "grad_norm": 2.9451184272766113, "learning_rate": 1.8068003983496943e-06, "loss": 0.0237, "step": 13080 }, { "epoch": 16.760563380281692, "grad_norm": 2.968151569366455, "learning_rate": 1.7996870109546167e-06, "loss": 0.0093, "step": 13090 }, { "epoch": 16.77336747759283, "grad_norm": 5.123749732971191, "learning_rate": 1.7925736235595393e-06, "loss": 0.0191, "step": 13100 }, { "epoch": 16.786171574903967, "grad_norm": 0.8735834360122681, "learning_rate": 1.7854602361644617e-06, "loss": 0.0216, "step": 13110 }, { "epoch": 16.798975672215107, "grad_norm": 1.7747591733932495, "learning_rate": 1.7783468487693843e-06, "loss": 0.0252, "step": 13120 }, { "epoch": 16.811779769526247, "grad_norm": 1.3483942747116089, "learning_rate": 1.7712334613743064e-06, "loss": 0.0209, "step": 13130 }, { "epoch": 16.824583866837386, "grad_norm": 0.00010964412649627775, "learning_rate": 1.7641200739792292e-06, "loss": 0.0269, "step": 13140 }, { "epoch": 16.837387964148526, "grad_norm": 2.804731845855713, "learning_rate": 1.7570066865841514e-06, "loss": 0.0163, "step": 13150 }, { "epoch": 16.850192061459666, "grad_norm": 8.251132011413574, "learning_rate": 1.7498932991890742e-06, "loss": 0.0318, "step": 13160 }, { "epoch": 16.862996158770805, "grad_norm": 7.249486923217773, "learning_rate": 1.7427799117939963e-06, "loss": 0.025, "step": 13170 }, { "epoch": 16.875800256081945, "grad_norm": 0.00013436282461043447, "learning_rate": 1.7356665243989191e-06, "loss": 0.0137, "step": 13180 }, { "epoch": 16.888604353393085, "grad_norm": 1.2946642637252808, "learning_rate": 1.7285531370038413e-06, "loss": 0.0177, "step": 13190 }, { "epoch": 16.901408450704224, "grad_norm": 5.95994758605957, "learning_rate": 1.7214397496087637e-06, "loss": 0.0176, "step": 13200 }, { "epoch": 16.914212548015364, "grad_norm": 0.05500904098153114, "learning_rate": 1.7143263622136863e-06, "loss": 0.0049, "step": 13210 }, { "epoch": 16.927016645326503, "grad_norm": 0.09497774392366409, "learning_rate": 1.7072129748186086e-06, "loss": 0.0239, "step": 13220 }, { "epoch": 16.939820742637643, "grad_norm": 4.39076566696167, "learning_rate": 1.7000995874235312e-06, "loss": 0.022, "step": 13230 }, { "epoch": 16.952624839948783, "grad_norm": 5.674398422241211, "learning_rate": 1.6929862000284536e-06, "loss": 0.0201, "step": 13240 }, { "epoch": 16.965428937259922, "grad_norm": 0.0006129290559329093, "learning_rate": 1.6858728126333762e-06, "loss": 0.0083, "step": 13250 }, { "epoch": 16.978233034571062, "grad_norm": 6.0583062171936035, "learning_rate": 1.6787594252382985e-06, "loss": 0.0269, "step": 13260 }, { "epoch": 16.9910371318822, "grad_norm": 1.6159683465957642, "learning_rate": 1.6716460378432211e-06, "loss": 0.0426, "step": 13270 }, { "epoch": 17.00384122919334, "grad_norm": 0.8002333641052246, "learning_rate": 1.6645326504481435e-06, "loss": 0.004, "step": 13280 }, { "epoch": 17.01664532650448, "grad_norm": 0.09969346225261688, "learning_rate": 1.657419263053066e-06, "loss": 0.0122, "step": 13290 }, { "epoch": 17.02944942381562, "grad_norm": 7.231577396392822, "learning_rate": 1.6503058756579885e-06, "loss": 0.0181, "step": 13300 }, { "epoch": 17.04225352112676, "grad_norm": 4.179574966430664, "learning_rate": 1.643192488262911e-06, "loss": 0.0212, "step": 13310 }, { "epoch": 17.0550576184379, "grad_norm": 2.9365389347076416, "learning_rate": 1.6360791008678334e-06, "loss": 0.0024, "step": 13320 }, { "epoch": 17.06786171574904, "grad_norm": 2.804448127746582, "learning_rate": 1.6289657134727558e-06, "loss": 0.0148, "step": 13330 }, { "epoch": 17.08066581306018, "grad_norm": 5.326770782470703, "learning_rate": 1.6218523260776784e-06, "loss": 0.0165, "step": 13340 }, { "epoch": 17.09346991037132, "grad_norm": 0.06416559964418411, "learning_rate": 1.6147389386826007e-06, "loss": 0.0094, "step": 13350 }, { "epoch": 17.106274007682458, "grad_norm": 1.2685847282409668, "learning_rate": 1.6076255512875233e-06, "loss": 0.019, "step": 13360 }, { "epoch": 17.119078104993598, "grad_norm": 2.1705806255340576, "learning_rate": 1.6005121638924457e-06, "loss": 0.0175, "step": 13370 }, { "epoch": 17.131882202304737, "grad_norm": 8.308648109436035, "learning_rate": 1.5933987764973683e-06, "loss": 0.0066, "step": 13380 }, { "epoch": 17.144686299615877, "grad_norm": 0.008040210232138634, "learning_rate": 1.5862853891022907e-06, "loss": 0.0136, "step": 13390 }, { "epoch": 17.157490396927017, "grad_norm": 0.00027485250029712915, "learning_rate": 1.5791720017072132e-06, "loss": 0.0212, "step": 13400 }, { "epoch": 17.170294494238156, "grad_norm": 2.500539541244507, "learning_rate": 1.5720586143121356e-06, "loss": 0.0073, "step": 13410 }, { "epoch": 17.183098591549296, "grad_norm": 3.578758716583252, "learning_rate": 1.5649452269170582e-06, "loss": 0.0237, "step": 13420 }, { "epoch": 17.195902688860436, "grad_norm": 0.06662648916244507, "learning_rate": 1.5578318395219804e-06, "loss": 0.0215, "step": 13430 }, { "epoch": 17.208706786171575, "grad_norm": 0.003631332889199257, "learning_rate": 1.5507184521269032e-06, "loss": 0.011, "step": 13440 }, { "epoch": 17.221510883482715, "grad_norm": 0.0027026438619941473, "learning_rate": 1.5436050647318253e-06, "loss": 0.0038, "step": 13450 }, { "epoch": 17.234314980793854, "grad_norm": 0.005422329064458609, "learning_rate": 1.5364916773367477e-06, "loss": 0.0016, "step": 13460 }, { "epoch": 17.247119078104994, "grad_norm": 0.016576915979385376, "learning_rate": 1.5293782899416703e-06, "loss": 0.0031, "step": 13470 }, { "epoch": 17.259923175416134, "grad_norm": 3.4552650451660156, "learning_rate": 1.5222649025465927e-06, "loss": 0.0372, "step": 13480 }, { "epoch": 17.272727272727273, "grad_norm": 2.6802520751953125, "learning_rate": 1.5151515151515152e-06, "loss": 0.0024, "step": 13490 }, { "epoch": 17.285531370038413, "grad_norm": 0.0830414816737175, "learning_rate": 1.5080381277564376e-06, "loss": 0.0113, "step": 13500 }, { "epoch": 17.298335467349553, "grad_norm": 1.7600657939910889, "learning_rate": 1.5009247403613602e-06, "loss": 0.0014, "step": 13510 }, { "epoch": 17.311139564660692, "grad_norm": 0.0025298339314758778, "learning_rate": 1.4938113529662826e-06, "loss": 0.0167, "step": 13520 }, { "epoch": 17.323943661971832, "grad_norm": 0.02072306163609028, "learning_rate": 1.4866979655712052e-06, "loss": 0.0095, "step": 13530 }, { "epoch": 17.33674775928297, "grad_norm": 6.340484142303467, "learning_rate": 1.4795845781761275e-06, "loss": 0.0453, "step": 13540 }, { "epoch": 17.34955185659411, "grad_norm": 0.010577944107353687, "learning_rate": 1.4724711907810501e-06, "loss": 0.0198, "step": 13550 }, { "epoch": 17.36235595390525, "grad_norm": 0.05272235721349716, "learning_rate": 1.4653578033859725e-06, "loss": 0.0279, "step": 13560 }, { "epoch": 17.37516005121639, "grad_norm": 1.1069378852844238, "learning_rate": 1.458244415990895e-06, "loss": 0.0112, "step": 13570 }, { "epoch": 17.38796414852753, "grad_norm": 0.24319998919963837, "learning_rate": 1.4511310285958174e-06, "loss": 0.0257, "step": 13580 }, { "epoch": 17.40076824583867, "grad_norm": 16.541982650756836, "learning_rate": 1.4440176412007398e-06, "loss": 0.0164, "step": 13590 }, { "epoch": 17.41357234314981, "grad_norm": 0.0016964910319074988, "learning_rate": 1.4369042538056624e-06, "loss": 0.0154, "step": 13600 }, { "epoch": 17.42637644046095, "grad_norm": 10.578028678894043, "learning_rate": 1.4297908664105848e-06, "loss": 0.0178, "step": 13610 }, { "epoch": 17.43918053777209, "grad_norm": 0.05916006490588188, "learning_rate": 1.4226774790155074e-06, "loss": 0.021, "step": 13620 }, { "epoch": 17.451984635083228, "grad_norm": 0.0031945924274623394, "learning_rate": 1.4155640916204297e-06, "loss": 0.0071, "step": 13630 }, { "epoch": 17.464788732394368, "grad_norm": 0.023241791874170303, "learning_rate": 1.4084507042253523e-06, "loss": 0.0152, "step": 13640 }, { "epoch": 17.477592829705507, "grad_norm": 0.058077480643987656, "learning_rate": 1.4013373168302747e-06, "loss": 0.003, "step": 13650 }, { "epoch": 17.490396927016647, "grad_norm": 10.599968910217285, "learning_rate": 1.3942239294351973e-06, "loss": 0.0476, "step": 13660 }, { "epoch": 17.503201024327787, "grad_norm": 6.625856399536133, "learning_rate": 1.3871105420401197e-06, "loss": 0.0325, "step": 13670 }, { "epoch": 17.516005121638923, "grad_norm": 5.916950225830078, "learning_rate": 1.3799971546450422e-06, "loss": 0.0194, "step": 13680 }, { "epoch": 17.528809218950062, "grad_norm": 0.0005128725315444171, "learning_rate": 1.3728837672499646e-06, "loss": 0.0273, "step": 13690 }, { "epoch": 17.541613316261202, "grad_norm": 0.0011182971065863967, "learning_rate": 1.3657703798548872e-06, "loss": 0.0279, "step": 13700 }, { "epoch": 17.55441741357234, "grad_norm": 7.456085681915283, "learning_rate": 1.3586569924598096e-06, "loss": 0.0312, "step": 13710 }, { "epoch": 17.56722151088348, "grad_norm": 0.41295695304870605, "learning_rate": 1.3515436050647317e-06, "loss": 0.0292, "step": 13720 }, { "epoch": 17.58002560819462, "grad_norm": 3.7234203815460205, "learning_rate": 1.3444302176696545e-06, "loss": 0.0031, "step": 13730 }, { "epoch": 17.59282970550576, "grad_norm": 0.37735211849212646, "learning_rate": 1.3373168302745767e-06, "loss": 0.0073, "step": 13740 }, { "epoch": 17.6056338028169, "grad_norm": 5.589832782745361, "learning_rate": 1.3302034428794993e-06, "loss": 0.0235, "step": 13750 }, { "epoch": 17.61843790012804, "grad_norm": 0.4297298192977905, "learning_rate": 1.3230900554844216e-06, "loss": 0.0211, "step": 13760 }, { "epoch": 17.63124199743918, "grad_norm": 1.479553461074829, "learning_rate": 1.3159766680893442e-06, "loss": 0.0282, "step": 13770 }, { "epoch": 17.64404609475032, "grad_norm": 0.000399995333282277, "learning_rate": 1.3088632806942666e-06, "loss": 0.0094, "step": 13780 }, { "epoch": 17.65685019206146, "grad_norm": 6.327033042907715, "learning_rate": 1.3017498932991892e-06, "loss": 0.009, "step": 13790 }, { "epoch": 17.6696542893726, "grad_norm": 5.372107982635498, "learning_rate": 1.2946365059041116e-06, "loss": 0.0098, "step": 13800 }, { "epoch": 17.682458386683738, "grad_norm": 1.6779136785771698e-05, "learning_rate": 1.2875231185090341e-06, "loss": 0.0265, "step": 13810 }, { "epoch": 17.695262483994878, "grad_norm": 0.167639821767807, "learning_rate": 1.2804097311139565e-06, "loss": 0.0104, "step": 13820 }, { "epoch": 17.708066581306017, "grad_norm": 0.00017483836563769728, "learning_rate": 1.2732963437188791e-06, "loss": 0.0128, "step": 13830 }, { "epoch": 17.720870678617157, "grad_norm": 0.05920706316828728, "learning_rate": 1.2661829563238015e-06, "loss": 0.0335, "step": 13840 }, { "epoch": 17.733674775928296, "grad_norm": 0.11115208268165588, "learning_rate": 1.2590695689287239e-06, "loss": 0.022, "step": 13850 }, { "epoch": 17.746478873239436, "grad_norm": 2.417051315307617, "learning_rate": 1.2519561815336464e-06, "loss": 0.0141, "step": 13860 }, { "epoch": 17.759282970550576, "grad_norm": 1.4494980573654175, "learning_rate": 1.244842794138569e-06, "loss": 0.0251, "step": 13870 }, { "epoch": 17.772087067861715, "grad_norm": 5.581284046173096, "learning_rate": 1.2377294067434914e-06, "loss": 0.0127, "step": 13880 }, { "epoch": 17.784891165172855, "grad_norm": 5.4958295822143555, "learning_rate": 1.2306160193484138e-06, "loss": 0.0292, "step": 13890 }, { "epoch": 17.797695262483995, "grad_norm": 7.042484283447266, "learning_rate": 1.2235026319533361e-06, "loss": 0.0177, "step": 13900 }, { "epoch": 17.810499359795134, "grad_norm": 0.0019001105101779103, "learning_rate": 1.2163892445582587e-06, "loss": 0.0136, "step": 13910 }, { "epoch": 17.823303457106274, "grad_norm": 0.005181374028325081, "learning_rate": 1.209275857163181e-06, "loss": 0.0065, "step": 13920 }, { "epoch": 17.836107554417413, "grad_norm": 0.0006981359329074621, "learning_rate": 1.2021624697681037e-06, "loss": 0.0042, "step": 13930 }, { "epoch": 17.848911651728553, "grad_norm": 3.4157283306121826, "learning_rate": 1.195049082373026e-06, "loss": 0.0239, "step": 13940 }, { "epoch": 17.861715749039693, "grad_norm": 0.0042902445420622826, "learning_rate": 1.1879356949779486e-06, "loss": 0.0084, "step": 13950 }, { "epoch": 17.874519846350832, "grad_norm": 10.998019218444824, "learning_rate": 1.180822307582871e-06, "loss": 0.0381, "step": 13960 }, { "epoch": 17.887323943661972, "grad_norm": 0.25908491015434265, "learning_rate": 1.1737089201877936e-06, "loss": 0.025, "step": 13970 }, { "epoch": 17.90012804097311, "grad_norm": 0.00100498937536031, "learning_rate": 1.166595532792716e-06, "loss": 0.0103, "step": 13980 }, { "epoch": 17.91293213828425, "grad_norm": 0.0007961208466440439, "learning_rate": 1.1594821453976386e-06, "loss": 0.006, "step": 13990 }, { "epoch": 17.92573623559539, "grad_norm": 0.3022013008594513, "learning_rate": 1.152368758002561e-06, "loss": 0.0012, "step": 14000 }, { "epoch": 17.93854033290653, "grad_norm": 0.2116912454366684, "learning_rate": 1.1452553706074835e-06, "loss": 0.0212, "step": 14010 }, { "epoch": 17.95134443021767, "grad_norm": 3.5195729732513428, "learning_rate": 1.1381419832124059e-06, "loss": 0.0137, "step": 14020 }, { "epoch": 17.96414852752881, "grad_norm": 0.3978916108608246, "learning_rate": 1.1310285958173283e-06, "loss": 0.013, "step": 14030 }, { "epoch": 17.97695262483995, "grad_norm": 0.5692456364631653, "learning_rate": 1.1239152084222506e-06, "loss": 0.0335, "step": 14040 }, { "epoch": 17.98975672215109, "grad_norm": 5.769397258758545, "learning_rate": 1.1168018210271732e-06, "loss": 0.0289, "step": 14050 }, { "epoch": 18.00256081946223, "grad_norm": 6.372897148132324, "learning_rate": 1.1096884336320956e-06, "loss": 0.0068, "step": 14060 }, { "epoch": 18.01536491677337, "grad_norm": 0.00024237189791165292, "learning_rate": 1.1025750462370182e-06, "loss": 0.0203, "step": 14070 }, { "epoch": 18.028169014084508, "grad_norm": 0.12732426822185516, "learning_rate": 1.0954616588419406e-06, "loss": 0.006, "step": 14080 }, { "epoch": 18.040973111395648, "grad_norm": 0.00020043789118062705, "learning_rate": 1.0883482714468631e-06, "loss": 0.0215, "step": 14090 }, { "epoch": 18.053777208706787, "grad_norm": 0.0010736729018390179, "learning_rate": 1.0812348840517855e-06, "loss": 0.0074, "step": 14100 }, { "epoch": 18.066581306017927, "grad_norm": 1.2515732049942017, "learning_rate": 1.074121496656708e-06, "loss": 0.0021, "step": 14110 }, { "epoch": 18.079385403329066, "grad_norm": 1.7571865320205688, "learning_rate": 1.0670081092616305e-06, "loss": 0.0029, "step": 14120 }, { "epoch": 18.092189500640206, "grad_norm": 0.773719072341919, "learning_rate": 1.059894721866553e-06, "loss": 0.0193, "step": 14130 }, { "epoch": 18.104993597951346, "grad_norm": 1.4211056232452393, "learning_rate": 1.0527813344714754e-06, "loss": 0.0072, "step": 14140 }, { "epoch": 18.117797695262485, "grad_norm": 8.346291542053223, "learning_rate": 1.045667947076398e-06, "loss": 0.0235, "step": 14150 }, { "epoch": 18.130601792573625, "grad_norm": 1.9946221113204956, "learning_rate": 1.0385545596813202e-06, "loss": 0.0048, "step": 14160 }, { "epoch": 18.143405889884765, "grad_norm": 0.26855456829071045, "learning_rate": 1.0314411722862428e-06, "loss": 0.0119, "step": 14170 }, { "epoch": 18.156209987195904, "grad_norm": 3.3399055004119873, "learning_rate": 1.0243277848911651e-06, "loss": 0.0103, "step": 14180 }, { "epoch": 18.169014084507044, "grad_norm": 4.981878757476807, "learning_rate": 1.0172143974960877e-06, "loss": 0.0176, "step": 14190 }, { "epoch": 18.181818181818183, "grad_norm": 0.00026743533089756966, "learning_rate": 1.01010101010101e-06, "loss": 0.0034, "step": 14200 }, { "epoch": 18.194622279129323, "grad_norm": 0.0003128933603875339, "learning_rate": 1.0029876227059327e-06, "loss": 0.0078, "step": 14210 }, { "epoch": 18.207426376440463, "grad_norm": 0.00414335448294878, "learning_rate": 9.95874235310855e-07, "loss": 0.0259, "step": 14220 }, { "epoch": 18.2202304737516, "grad_norm": 0.105068139731884, "learning_rate": 9.887608479157776e-07, "loss": 0.0069, "step": 14230 }, { "epoch": 18.23303457106274, "grad_norm": 7.0577216148376465, "learning_rate": 9.816474605207e-07, "loss": 0.0187, "step": 14240 }, { "epoch": 18.245838668373878, "grad_norm": 0.00019036579760722816, "learning_rate": 9.745340731256226e-07, "loss": 0.0162, "step": 14250 }, { "epoch": 18.258642765685018, "grad_norm": 10.219581604003906, "learning_rate": 9.67420685730545e-07, "loss": 0.0144, "step": 14260 }, { "epoch": 18.271446862996157, "grad_norm": 7.480413436889648, "learning_rate": 9.603072983354676e-07, "loss": 0.0192, "step": 14270 }, { "epoch": 18.284250960307297, "grad_norm": 5.9122114181518555, "learning_rate": 9.531939109403899e-07, "loss": 0.0079, "step": 14280 }, { "epoch": 18.297055057618437, "grad_norm": 9.46755599975586, "learning_rate": 9.460805235453123e-07, "loss": 0.0258, "step": 14290 }, { "epoch": 18.309859154929576, "grad_norm": 7.41144049243303e-06, "learning_rate": 9.389671361502348e-07, "loss": 0.0186, "step": 14300 }, { "epoch": 18.322663252240716, "grad_norm": 0.0027089251670986414, "learning_rate": 9.318537487551573e-07, "loss": 0.0162, "step": 14310 }, { "epoch": 18.335467349551855, "grad_norm": 0.007182203698903322, "learning_rate": 9.247403613600797e-07, "loss": 0.0148, "step": 14320 }, { "epoch": 18.348271446862995, "grad_norm": 0.0003917156718671322, "learning_rate": 9.176269739650022e-07, "loss": 0.0136, "step": 14330 }, { "epoch": 18.361075544174135, "grad_norm": 0.00159465370234102, "learning_rate": 9.105135865699247e-07, "loss": 0.0003, "step": 14340 }, { "epoch": 18.373879641485274, "grad_norm": 3.8067879676818848, "learning_rate": 9.034001991748472e-07, "loss": 0.0108, "step": 14350 }, { "epoch": 18.386683738796414, "grad_norm": 1.2124892473220825, "learning_rate": 8.962868117797697e-07, "loss": 0.0088, "step": 14360 }, { "epoch": 18.399487836107554, "grad_norm": 5.614289283752441, "learning_rate": 8.891734243846921e-07, "loss": 0.0107, "step": 14370 }, { "epoch": 18.412291933418693, "grad_norm": 8.393487930297852, "learning_rate": 8.820600369896146e-07, "loss": 0.0262, "step": 14380 }, { "epoch": 18.425096030729833, "grad_norm": 0.03645249456167221, "learning_rate": 8.749466495945371e-07, "loss": 0.0367, "step": 14390 }, { "epoch": 18.437900128040972, "grad_norm": 0.003539648372679949, "learning_rate": 8.678332621994596e-07, "loss": 0.0224, "step": 14400 }, { "epoch": 18.450704225352112, "grad_norm": 0.6319210529327393, "learning_rate": 8.607198748043818e-07, "loss": 0.0136, "step": 14410 }, { "epoch": 18.46350832266325, "grad_norm": 0.0015524758491665125, "learning_rate": 8.536064874093043e-07, "loss": 0.0056, "step": 14420 }, { "epoch": 18.47631241997439, "grad_norm": 0.022453349083662033, "learning_rate": 8.464931000142268e-07, "loss": 0.0345, "step": 14430 }, { "epoch": 18.48911651728553, "grad_norm": 0.00012319152301643044, "learning_rate": 8.393797126191493e-07, "loss": 0.0035, "step": 14440 }, { "epoch": 18.50192061459667, "grad_norm": 0.07099111378192902, "learning_rate": 8.322663252240718e-07, "loss": 0.0054, "step": 14450 }, { "epoch": 18.51472471190781, "grad_norm": 0.3618342876434326, "learning_rate": 8.251529378289942e-07, "loss": 0.028, "step": 14460 }, { "epoch": 18.52752880921895, "grad_norm": 0.0001998326915781945, "learning_rate": 8.180395504339167e-07, "loss": 0.0342, "step": 14470 }, { "epoch": 18.54033290653009, "grad_norm": 11.499155044555664, "learning_rate": 8.109261630388392e-07, "loss": 0.0276, "step": 14480 }, { "epoch": 18.55313700384123, "grad_norm": 0.010596271604299545, "learning_rate": 8.038127756437617e-07, "loss": 0.0141, "step": 14490 }, { "epoch": 18.56594110115237, "grad_norm": 0.042814165353775024, "learning_rate": 7.966993882486841e-07, "loss": 0.0114, "step": 14500 }, { "epoch": 18.57874519846351, "grad_norm": 0.9915781617164612, "learning_rate": 7.895860008536066e-07, "loss": 0.0166, "step": 14510 }, { "epoch": 18.591549295774648, "grad_norm": 0.3098224699497223, "learning_rate": 7.824726134585291e-07, "loss": 0.0221, "step": 14520 }, { "epoch": 18.604353393085788, "grad_norm": 0.7572695016860962, "learning_rate": 7.753592260634516e-07, "loss": 0.0077, "step": 14530 }, { "epoch": 18.617157490396927, "grad_norm": 7.673007965087891, "learning_rate": 7.682458386683739e-07, "loss": 0.0224, "step": 14540 }, { "epoch": 18.629961587708067, "grad_norm": 0.04567665979266167, "learning_rate": 7.611324512732963e-07, "loss": 0.025, "step": 14550 }, { "epoch": 18.642765685019207, "grad_norm": 4.875103950500488, "learning_rate": 7.540190638782188e-07, "loss": 0.0065, "step": 14560 }, { "epoch": 18.655569782330346, "grad_norm": 5.814393997192383, "learning_rate": 7.469056764831413e-07, "loss": 0.0183, "step": 14570 }, { "epoch": 18.668373879641486, "grad_norm": 0.002308470429852605, "learning_rate": 7.397922890880638e-07, "loss": 0.0101, "step": 14580 }, { "epoch": 18.681177976952625, "grad_norm": 0.00015365192666649818, "learning_rate": 7.326789016929862e-07, "loss": 0.0062, "step": 14590 }, { "epoch": 18.693982074263765, "grad_norm": 0.02845505252480507, "learning_rate": 7.255655142979087e-07, "loss": 0.0085, "step": 14600 }, { "epoch": 18.706786171574905, "grad_norm": 0.01130868960171938, "learning_rate": 7.184521269028312e-07, "loss": 0.0074, "step": 14610 }, { "epoch": 18.719590268886044, "grad_norm": 6.3299174308776855, "learning_rate": 7.113387395077537e-07, "loss": 0.0115, "step": 14620 }, { "epoch": 18.732394366197184, "grad_norm": 1.7763944864273071, "learning_rate": 7.042253521126762e-07, "loss": 0.0186, "step": 14630 }, { "epoch": 18.745198463508324, "grad_norm": 5.388847351074219, "learning_rate": 6.971119647175986e-07, "loss": 0.0316, "step": 14640 }, { "epoch": 18.758002560819463, "grad_norm": 1.5063540935516357, "learning_rate": 6.899985773225211e-07, "loss": 0.0086, "step": 14650 }, { "epoch": 18.770806658130603, "grad_norm": 3.6523196697235107, "learning_rate": 6.828851899274436e-07, "loss": 0.0126, "step": 14660 }, { "epoch": 18.783610755441742, "grad_norm": 2.449568271636963, "learning_rate": 6.757718025323659e-07, "loss": 0.0369, "step": 14670 }, { "epoch": 18.796414852752882, "grad_norm": 0.005643006414175034, "learning_rate": 6.686584151372883e-07, "loss": 0.0061, "step": 14680 }, { "epoch": 18.80921895006402, "grad_norm": 5.944687843322754, "learning_rate": 6.615450277422108e-07, "loss": 0.0095, "step": 14690 }, { "epoch": 18.82202304737516, "grad_norm": 0.0013825156493112445, "learning_rate": 6.544316403471333e-07, "loss": 0.0117, "step": 14700 }, { "epoch": 18.8348271446863, "grad_norm": 3.3072972297668457, "learning_rate": 6.473182529520558e-07, "loss": 0.0125, "step": 14710 }, { "epoch": 18.84763124199744, "grad_norm": 0.058027319610118866, "learning_rate": 6.402048655569783e-07, "loss": 0.0233, "step": 14720 }, { "epoch": 18.86043533930858, "grad_norm": 4.892437934875488, "learning_rate": 6.330914781619007e-07, "loss": 0.0141, "step": 14730 }, { "epoch": 18.87323943661972, "grad_norm": 1.4201620817184448, "learning_rate": 6.259780907668232e-07, "loss": 0.0122, "step": 14740 }, { "epoch": 18.88604353393086, "grad_norm": 0.0010934326564893126, "learning_rate": 6.188647033717457e-07, "loss": 0.027, "step": 14750 }, { "epoch": 18.898847631242, "grad_norm": 8.750072479248047, "learning_rate": 6.117513159766681e-07, "loss": 0.0148, "step": 14760 }, { "epoch": 18.91165172855314, "grad_norm": 17.716455459594727, "learning_rate": 6.046379285815906e-07, "loss": 0.0458, "step": 14770 }, { "epoch": 18.92445582586428, "grad_norm": 0.10776586830615997, "learning_rate": 5.97524541186513e-07, "loss": 0.0123, "step": 14780 }, { "epoch": 18.937259923175418, "grad_norm": 3.1260874271392822, "learning_rate": 5.904111537914355e-07, "loss": 0.0041, "step": 14790 }, { "epoch": 18.950064020486554, "grad_norm": 0.0927921012043953, "learning_rate": 5.83297766396358e-07, "loss": 0.0279, "step": 14800 }, { "epoch": 18.962868117797694, "grad_norm": 0.2085677683353424, "learning_rate": 5.761843790012805e-07, "loss": 0.0251, "step": 14810 }, { "epoch": 18.975672215108833, "grad_norm": 0.1508549600839615, "learning_rate": 5.690709916062029e-07, "loss": 0.0114, "step": 14820 }, { "epoch": 18.988476312419973, "grad_norm": 0.857793927192688, "learning_rate": 5.619576042111253e-07, "loss": 0.0049, "step": 14830 }, { "epoch": 19.001280409731113, "grad_norm": 0.002058560959994793, "learning_rate": 5.548442168160478e-07, "loss": 0.0219, "step": 14840 }, { "epoch": 19.014084507042252, "grad_norm": 0.0020185494795441628, "learning_rate": 5.477308294209703e-07, "loss": 0.0096, "step": 14850 }, { "epoch": 19.026888604353392, "grad_norm": 0.00022442065528593957, "learning_rate": 5.406174420258928e-07, "loss": 0.0068, "step": 14860 }, { "epoch": 19.03969270166453, "grad_norm": 0.17310723662376404, "learning_rate": 5.335040546308152e-07, "loss": 0.0063, "step": 14870 }, { "epoch": 19.05249679897567, "grad_norm": 0.008212696760892868, "learning_rate": 5.263906672357377e-07, "loss": 0.0112, "step": 14880 }, { "epoch": 19.06530089628681, "grad_norm": 2.785284596029669e-05, "learning_rate": 5.192772798406601e-07, "loss": 0.0218, "step": 14890 }, { "epoch": 19.07810499359795, "grad_norm": 0.023113705217838287, "learning_rate": 5.121638924455826e-07, "loss": 0.0286, "step": 14900 }, { "epoch": 19.09090909090909, "grad_norm": 0.005509368143975735, "learning_rate": 5.05050505050505e-07, "loss": 0.0143, "step": 14910 }, { "epoch": 19.10371318822023, "grad_norm": 0.008240880444645882, "learning_rate": 4.979371176554275e-07, "loss": 0.0054, "step": 14920 }, { "epoch": 19.11651728553137, "grad_norm": 1.2594153881072998, "learning_rate": 4.9082373026035e-07, "loss": 0.0036, "step": 14930 }, { "epoch": 19.12932138284251, "grad_norm": 4.33413553237915, "learning_rate": 4.837103428652725e-07, "loss": 0.0092, "step": 14940 }, { "epoch": 19.14212548015365, "grad_norm": 2.718404531478882, "learning_rate": 4.7659695547019496e-07, "loss": 0.0068, "step": 14950 }, { "epoch": 19.154929577464788, "grad_norm": 0.0012664749519899487, "learning_rate": 4.694835680751174e-07, "loss": 0.0096, "step": 14960 }, { "epoch": 19.167733674775928, "grad_norm": 0.0038974243216216564, "learning_rate": 4.6237018068003987e-07, "loss": 0.0069, "step": 14970 }, { "epoch": 19.180537772087067, "grad_norm": 7.048045635223389, "learning_rate": 4.5525679328496235e-07, "loss": 0.0145, "step": 14980 }, { "epoch": 19.193341869398207, "grad_norm": 5.328739166259766, "learning_rate": 4.481434058898848e-07, "loss": 0.0215, "step": 14990 }, { "epoch": 19.206145966709347, "grad_norm": 0.35929474234580994, "learning_rate": 4.410300184948073e-07, "loss": 0.0327, "step": 15000 }, { "epoch": 19.218950064020486, "grad_norm": 5.503216743469238, "learning_rate": 4.339166310997298e-07, "loss": 0.0231, "step": 15010 }, { "epoch": 19.231754161331626, "grad_norm": 0.07661139219999313, "learning_rate": 4.2680324370465216e-07, "loss": 0.0245, "step": 15020 }, { "epoch": 19.244558258642765, "grad_norm": 0.8475786447525024, "learning_rate": 4.1968985630957464e-07, "loss": 0.0248, "step": 15030 }, { "epoch": 19.257362355953905, "grad_norm": 0.0011420133523643017, "learning_rate": 4.125764689144971e-07, "loss": 0.0178, "step": 15040 }, { "epoch": 19.270166453265045, "grad_norm": 0.0012623609509319067, "learning_rate": 4.054630815194196e-07, "loss": 0.0154, "step": 15050 }, { "epoch": 19.282970550576184, "grad_norm": 0.053852129727602005, "learning_rate": 3.9834969412434207e-07, "loss": 0.0264, "step": 15060 }, { "epoch": 19.295774647887324, "grad_norm": 0.0035585970617830753, "learning_rate": 3.9123630672926455e-07, "loss": 0.0044, "step": 15070 }, { "epoch": 19.308578745198464, "grad_norm": 5.149901390075684, "learning_rate": 3.841229193341869e-07, "loss": 0.0212, "step": 15080 }, { "epoch": 19.321382842509603, "grad_norm": 0.03466752544045448, "learning_rate": 3.770095319391094e-07, "loss": 0.0033, "step": 15090 }, { "epoch": 19.334186939820743, "grad_norm": 0.0002815356128849089, "learning_rate": 3.698961445440319e-07, "loss": 0.0091, "step": 15100 }, { "epoch": 19.346991037131883, "grad_norm": 0.037533290684223175, "learning_rate": 3.6278275714895436e-07, "loss": 0.0159, "step": 15110 }, { "epoch": 19.359795134443022, "grad_norm": 0.12547887861728668, "learning_rate": 3.5566936975387684e-07, "loss": 0.0142, "step": 15120 }, { "epoch": 19.37259923175416, "grad_norm": 1.4191429615020752, "learning_rate": 3.485559823587993e-07, "loss": 0.0247, "step": 15130 }, { "epoch": 19.3854033290653, "grad_norm": 0.0016714661614969373, "learning_rate": 3.414425949637218e-07, "loss": 0.0067, "step": 15140 }, { "epoch": 19.39820742637644, "grad_norm": 0.08134725689888, "learning_rate": 3.3432920756864417e-07, "loss": 0.0286, "step": 15150 }, { "epoch": 19.41101152368758, "grad_norm": 0.005342427175492048, "learning_rate": 3.2721582017356665e-07, "loss": 0.0168, "step": 15160 }, { "epoch": 19.42381562099872, "grad_norm": 0.06078413128852844, "learning_rate": 3.2010243277848913e-07, "loss": 0.018, "step": 15170 }, { "epoch": 19.43661971830986, "grad_norm": 0.023634914308786392, "learning_rate": 3.129890453834116e-07, "loss": 0.0143, "step": 15180 }, { "epoch": 19.449423815621, "grad_norm": 9.950129508972168, "learning_rate": 3.0587565798833404e-07, "loss": 0.0089, "step": 15190 }, { "epoch": 19.46222791293214, "grad_norm": 5.060400485992432, "learning_rate": 2.987622705932565e-07, "loss": 0.0204, "step": 15200 }, { "epoch": 19.47503201024328, "grad_norm": 0.000995470443740487, "learning_rate": 2.91648883198179e-07, "loss": 0.023, "step": 15210 }, { "epoch": 19.48783610755442, "grad_norm": 0.0009162668720819056, "learning_rate": 2.8453549580310147e-07, "loss": 0.0043, "step": 15220 }, { "epoch": 19.500640204865558, "grad_norm": 0.0008124169544316828, "learning_rate": 2.774221084080239e-07, "loss": 0.0163, "step": 15230 }, { "epoch": 19.513444302176698, "grad_norm": 3.7467453479766846, "learning_rate": 2.703087210129464e-07, "loss": 0.0037, "step": 15240 }, { "epoch": 19.526248399487837, "grad_norm": 0.6744768023490906, "learning_rate": 2.6319533361786886e-07, "loss": 0.0129, "step": 15250 }, { "epoch": 19.539052496798977, "grad_norm": 0.02649999037384987, "learning_rate": 2.560819462227913e-07, "loss": 0.0338, "step": 15260 }, { "epoch": 19.551856594110117, "grad_norm": 0.5284630060195923, "learning_rate": 2.4896855882771376e-07, "loss": 0.0098, "step": 15270 }, { "epoch": 19.564660691421256, "grad_norm": 0.0007515403558500111, "learning_rate": 2.4185517143263624e-07, "loss": 0.0081, "step": 15280 }, { "epoch": 19.577464788732396, "grad_norm": 4.211970806121826, "learning_rate": 2.347417840375587e-07, "loss": 0.0122, "step": 15290 }, { "epoch": 19.590268886043535, "grad_norm": 0.0010327131021767855, "learning_rate": 2.2762839664248117e-07, "loss": 0.0149, "step": 15300 }, { "epoch": 19.603072983354675, "grad_norm": 0.00021798952366225421, "learning_rate": 2.2051500924740365e-07, "loss": 0.0042, "step": 15310 }, { "epoch": 19.615877080665815, "grad_norm": 7.364733695983887, "learning_rate": 2.1340162185232608e-07, "loss": 0.0177, "step": 15320 }, { "epoch": 19.628681177976954, "grad_norm": 0.023925766348838806, "learning_rate": 2.0628823445724856e-07, "loss": 0.0036, "step": 15330 }, { "epoch": 19.64148527528809, "grad_norm": 0.04516793414950371, "learning_rate": 1.9917484706217104e-07, "loss": 0.0316, "step": 15340 }, { "epoch": 19.65428937259923, "grad_norm": 7.666085243225098, "learning_rate": 1.9206145966709346e-07, "loss": 0.0301, "step": 15350 }, { "epoch": 19.66709346991037, "grad_norm": 2.243300676345825, "learning_rate": 1.8494807227201594e-07, "loss": 0.0057, "step": 15360 }, { "epoch": 19.67989756722151, "grad_norm": 0.002540783490985632, "learning_rate": 1.7783468487693842e-07, "loss": 0.0076, "step": 15370 }, { "epoch": 19.69270166453265, "grad_norm": 4.108165264129639, "learning_rate": 1.707212974818609e-07, "loss": 0.0129, "step": 15380 }, { "epoch": 19.70550576184379, "grad_norm": 0.00024948702775873244, "learning_rate": 1.6360791008678333e-07, "loss": 0.0339, "step": 15390 }, { "epoch": 19.718309859154928, "grad_norm": 0.04997668415307999, "learning_rate": 1.564945226917058e-07, "loss": 0.0144, "step": 15400 }, { "epoch": 19.731113956466068, "grad_norm": 0.7676925659179688, "learning_rate": 1.4938113529662826e-07, "loss": 0.0096, "step": 15410 }, { "epoch": 19.743918053777207, "grad_norm": 0.0016744857421144843, "learning_rate": 1.4226774790155074e-07, "loss": 0.0084, "step": 15420 }, { "epoch": 19.756722151088347, "grad_norm": 0.19571436941623688, "learning_rate": 1.351543605064732e-07, "loss": 0.0079, "step": 15430 }, { "epoch": 19.769526248399487, "grad_norm": 4.245015621185303, "learning_rate": 1.2804097311139564e-07, "loss": 0.0282, "step": 15440 }, { "epoch": 19.782330345710626, "grad_norm": 7.645667210454121e-05, "learning_rate": 1.2092758571631812e-07, "loss": 0.0155, "step": 15450 }, { "epoch": 19.795134443021766, "grad_norm": 0.0017080483958125114, "learning_rate": 1.1381419832124059e-07, "loss": 0.0077, "step": 15460 }, { "epoch": 19.807938540332906, "grad_norm": 10.036234855651855, "learning_rate": 1.0670081092616304e-07, "loss": 0.0109, "step": 15470 }, { "epoch": 19.820742637644045, "grad_norm": 12.06249713897705, "learning_rate": 9.958742353108552e-08, "loss": 0.0156, "step": 15480 }, { "epoch": 19.833546734955185, "grad_norm": 0.14088483154773712, "learning_rate": 9.247403613600797e-08, "loss": 0.0295, "step": 15490 }, { "epoch": 19.846350832266324, "grad_norm": 6.773251533508301, "learning_rate": 8.536064874093045e-08, "loss": 0.0338, "step": 15500 }, { "epoch": 19.859154929577464, "grad_norm": 0.03990509361028671, "learning_rate": 7.82472613458529e-08, "loss": 0.0138, "step": 15510 }, { "epoch": 19.871959026888604, "grad_norm": 13.610167503356934, "learning_rate": 7.113387395077537e-08, "loss": 0.0102, "step": 15520 }, { "epoch": 19.884763124199743, "grad_norm": 5.543415682041086e-05, "learning_rate": 6.402048655569782e-08, "loss": 0.0117, "step": 15530 }, { "epoch": 19.897567221510883, "grad_norm": 7.679758548736572, "learning_rate": 5.690709916062029e-08, "loss": 0.0101, "step": 15540 }, { "epoch": 19.910371318822023, "grad_norm": 0.32348430156707764, "learning_rate": 4.979371176554276e-08, "loss": 0.0097, "step": 15550 }, { "epoch": 19.923175416133162, "grad_norm": 0.0029495363123714924, "learning_rate": 4.2680324370465225e-08, "loss": 0.0119, "step": 15560 }, { "epoch": 19.935979513444302, "grad_norm": 3.074021339416504, "learning_rate": 3.5566936975387684e-08, "loss": 0.0116, "step": 15570 }, { "epoch": 19.94878361075544, "grad_norm": 0.3156638443470001, "learning_rate": 2.8453549580310147e-08, "loss": 0.0066, "step": 15580 }, { "epoch": 19.96158770806658, "grad_norm": 4.489070892333984, "learning_rate": 2.1340162185232612e-08, "loss": 0.009, "step": 15590 }, { "epoch": 19.97439180537772, "grad_norm": 1.1894117960764561e-05, "learning_rate": 1.4226774790155073e-08, "loss": 0.0101, "step": 15600 }, { "epoch": 19.98719590268886, "grad_norm": 0.0012836161768063903, "learning_rate": 7.113387395077537e-09, "loss": 0.0119, "step": 15610 }, { "epoch": 20.0, "grad_norm": 0.0017099727410823107, "learning_rate": 0.0, "loss": 0.0194, "step": 15620 } ], "logging_steps": 10, "max_steps": 15620, "num_input_tokens_seen": 0, "num_train_epochs": 20, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 64, "trial_name": null, "trial_params": null }