{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.7094017094017095, "eval_steps": 500, "global_step": 3600, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.004748338081671415, "grad_norm": 18.611501136425524, "learning_rate": 8.547008547008548e-07, "loss": 8.2433, "step": 10 }, { "epoch": 0.00949667616334283, "grad_norm": 17.51436529626854, "learning_rate": 1.8043684710351376e-06, "loss": 8.133, "step": 20 }, { "epoch": 0.014245014245014245, "grad_norm": 9.119462674296598, "learning_rate": 2.7540360873694207e-06, "loss": 7.5221, "step": 30 }, { "epoch": 0.01899335232668566, "grad_norm": 4.680965546530251, "learning_rate": 3.7037037037037037e-06, "loss": 6.8105, "step": 40 }, { "epoch": 0.023741690408357077, "grad_norm": 2.8843538584400394, "learning_rate": 4.653371320037987e-06, "loss": 6.3644, "step": 50 }, { "epoch": 0.02849002849002849, "grad_norm": 2.7183786105329797, "learning_rate": 5.60303893637227e-06, "loss": 6.0537, "step": 60 }, { "epoch": 0.03323836657169991, "grad_norm": 3.1505246290978097, "learning_rate": 6.5527065527065525e-06, "loss": 5.7138, "step": 70 }, { "epoch": 0.03798670465337132, "grad_norm": 3.3227597395164583, "learning_rate": 7.5023741690408365e-06, "loss": 5.2202, "step": 80 }, { "epoch": 0.042735042735042736, "grad_norm": 4.796158007024566, "learning_rate": 8.452041785375119e-06, "loss": 4.6939, "step": 90 }, { "epoch": 0.04748338081671415, "grad_norm": 4.22273266005884, "learning_rate": 9.401709401709402e-06, "loss": 4.2314, "step": 100 }, { "epoch": 0.052231718898385564, "grad_norm": 5.083224986055868, "learning_rate": 1.0351377018043685e-05, "loss": 3.8622, "step": 110 }, { "epoch": 0.05698005698005698, "grad_norm": 4.62630512353402, "learning_rate": 1.1301044634377968e-05, "loss": 3.5804, "step": 120 }, { "epoch": 0.06172839506172839, "grad_norm": 8.968159181123188, "learning_rate": 1.2250712250712251e-05, "loss": 3.3623, "step": 130 }, { "epoch": 0.06647673314339982, "grad_norm": 6.284759849451774, "learning_rate": 1.3200379867046534e-05, "loss": 3.208, "step": 140 }, { "epoch": 0.07122507122507123, "grad_norm": 7.873569595607114, "learning_rate": 1.4150047483380819e-05, "loss": 3.0847, "step": 150 }, { "epoch": 0.07597340930674264, "grad_norm": 7.5195257049383715, "learning_rate": 1.50997150997151e-05, "loss": 2.9765, "step": 160 }, { "epoch": 0.08072174738841406, "grad_norm": 10.085451011841057, "learning_rate": 1.604938271604938e-05, "loss": 2.891, "step": 170 }, { "epoch": 0.08547008547008547, "grad_norm": 7.574979735730398, "learning_rate": 1.6999050332383668e-05, "loss": 2.8186, "step": 180 }, { "epoch": 0.09021842355175688, "grad_norm": 10.874635360709647, "learning_rate": 1.794871794871795e-05, "loss": 2.7532, "step": 190 }, { "epoch": 0.0949667616334283, "grad_norm": 9.110042663641787, "learning_rate": 1.889838556505223e-05, "loss": 2.6907, "step": 200 }, { "epoch": 0.09971509971509972, "grad_norm": 10.675571155059778, "learning_rate": 1.9848053181386517e-05, "loss": 2.6273, "step": 210 }, { "epoch": 0.10446343779677113, "grad_norm": 10.836164458180809, "learning_rate": 2.07977207977208e-05, "loss": 2.5466, "step": 220 }, { "epoch": 0.10921177587844255, "grad_norm": 10.441632304833055, "learning_rate": 2.174738841405508e-05, "loss": 2.4771, "step": 230 }, { "epoch": 0.11396011396011396, "grad_norm": 10.652520415153692, "learning_rate": 2.2697056030389367e-05, "loss": 2.4495, "step": 240 }, { "epoch": 0.11870845204178537, "grad_norm": 14.205715643733297, "learning_rate": 2.364672364672365e-05, "loss": 2.3755, "step": 250 }, { "epoch": 0.12345679012345678, "grad_norm": 18.018386067888525, "learning_rate": 2.459639126305793e-05, "loss": 2.3259, "step": 260 }, { "epoch": 0.1282051282051282, "grad_norm": 11.148502639062643, "learning_rate": 2.5546058879392216e-05, "loss": 2.2627, "step": 270 }, { "epoch": 0.13295346628679963, "grad_norm": 16.45588748254632, "learning_rate": 2.64957264957265e-05, "loss": 2.1778, "step": 280 }, { "epoch": 0.13770180436847104, "grad_norm": 10.54821843220057, "learning_rate": 2.744539411206078e-05, "loss": 2.0719, "step": 290 }, { "epoch": 0.14245014245014245, "grad_norm": 15.431971282739063, "learning_rate": 2.839506172839506e-05, "loss": 1.9551, "step": 300 }, { "epoch": 0.14719848053181386, "grad_norm": 10.589892377462267, "learning_rate": 2.9344729344729345e-05, "loss": 1.8346, "step": 310 }, { "epoch": 0.15194681861348527, "grad_norm": 11.4410671979004, "learning_rate": 3.029439696106363e-05, "loss": 1.7393, "step": 320 }, { "epoch": 0.15669515669515668, "grad_norm": 8.801809080434843, "learning_rate": 3.1244064577397914e-05, "loss": 1.6355, "step": 330 }, { "epoch": 0.16144349477682812, "grad_norm": 11.031833382378403, "learning_rate": 3.2193732193732194e-05, "loss": 1.5394, "step": 340 }, { "epoch": 0.16619183285849953, "grad_norm": 10.880011823678203, "learning_rate": 3.314339981006648e-05, "loss": 1.4806, "step": 350 }, { "epoch": 0.17094017094017094, "grad_norm": 8.202709729567582, "learning_rate": 3.409306742640076e-05, "loss": 1.4215, "step": 360 }, { "epoch": 0.17568850902184235, "grad_norm": 6.835058852289836, "learning_rate": 3.504273504273504e-05, "loss": 1.3542, "step": 370 }, { "epoch": 0.18043684710351376, "grad_norm": 6.290833237211221, "learning_rate": 3.599240265906933e-05, "loss": 1.2924, "step": 380 }, { "epoch": 0.18518518518518517, "grad_norm": 6.373950257766728, "learning_rate": 3.694207027540361e-05, "loss": 1.246, "step": 390 }, { "epoch": 0.1899335232668566, "grad_norm": 5.138017574592672, "learning_rate": 3.789173789173789e-05, "loss": 1.2006, "step": 400 }, { "epoch": 0.19468186134852802, "grad_norm": 5.516477133552481, "learning_rate": 3.884140550807218e-05, "loss": 1.1474, "step": 410 }, { "epoch": 0.19943019943019943, "grad_norm": 9.044407794994493, "learning_rate": 3.979107312440646e-05, "loss": 1.1134, "step": 420 }, { "epoch": 0.20417853751187084, "grad_norm": 5.242982362793433, "learning_rate": 4.074074074074074e-05, "loss": 1.07, "step": 430 }, { "epoch": 0.20892687559354226, "grad_norm": 9.242048997048826, "learning_rate": 4.169040835707503e-05, "loss": 1.0556, "step": 440 }, { "epoch": 0.21367521367521367, "grad_norm": 6.622904919860144, "learning_rate": 4.264007597340931e-05, "loss": 1.0161, "step": 450 }, { "epoch": 0.2184235517568851, "grad_norm": 5.219976810802862, "learning_rate": 4.358974358974359e-05, "loss": 0.9858, "step": 460 }, { "epoch": 0.22317188983855651, "grad_norm": 7.359755852742898, "learning_rate": 4.453941120607788e-05, "loss": 0.9481, "step": 470 }, { "epoch": 0.22792022792022792, "grad_norm": 4.578299758997878, "learning_rate": 4.548907882241216e-05, "loss": 0.916, "step": 480 }, { "epoch": 0.23266856600189934, "grad_norm": 5.343119156721728, "learning_rate": 4.643874643874644e-05, "loss": 0.8943, "step": 490 }, { "epoch": 0.23741690408357075, "grad_norm": 5.4714925792960205, "learning_rate": 4.738841405508073e-05, "loss": 0.8707, "step": 500 }, { "epoch": 0.24216524216524216, "grad_norm": 5.0402481039237506, "learning_rate": 4.833808167141501e-05, "loss": 0.8533, "step": 510 }, { "epoch": 0.24691358024691357, "grad_norm": 5.05824568493684, "learning_rate": 4.928774928774929e-05, "loss": 0.8179, "step": 520 }, { "epoch": 0.251661918328585, "grad_norm": 5.072707834549211, "learning_rate": 5.0237416904083576e-05, "loss": 0.7882, "step": 530 }, { "epoch": 0.2564102564102564, "grad_norm": 4.566630961707184, "learning_rate": 5.118708452041786e-05, "loss": 0.7737, "step": 540 }, { "epoch": 0.2611585944919278, "grad_norm": 5.782147763909198, "learning_rate": 5.213675213675214e-05, "loss": 0.7555, "step": 550 }, { "epoch": 0.26590693257359926, "grad_norm": 6.5118194373773655, "learning_rate": 5.308641975308642e-05, "loss": 0.7249, "step": 560 }, { "epoch": 0.2706552706552707, "grad_norm": 5.823368294624547, "learning_rate": 5.403608736942071e-05, "loss": 0.7082, "step": 570 }, { "epoch": 0.2754036087369421, "grad_norm": 4.530039848879685, "learning_rate": 5.498575498575499e-05, "loss": 0.6823, "step": 580 }, { "epoch": 0.2801519468186135, "grad_norm": 4.248480448022572, "learning_rate": 5.5935422602089274e-05, "loss": 0.666, "step": 590 }, { "epoch": 0.2849002849002849, "grad_norm": 5.06442063116728, "learning_rate": 5.6885090218423554e-05, "loss": 0.6461, "step": 600 }, { "epoch": 0.2896486229819563, "grad_norm": 6.882191487258264, "learning_rate": 5.7834757834757834e-05, "loss": 0.6338, "step": 610 }, { "epoch": 0.2943969610636277, "grad_norm": 4.246010140777778, "learning_rate": 5.878442545109212e-05, "loss": 0.6021, "step": 620 }, { "epoch": 0.29914529914529914, "grad_norm": 5.977623056918047, "learning_rate": 5.97340930674264e-05, "loss": 0.5866, "step": 630 }, { "epoch": 0.30389363722697055, "grad_norm": 5.1933089523540525, "learning_rate": 6.068376068376068e-05, "loss": 0.5701, "step": 640 }, { "epoch": 0.30864197530864196, "grad_norm": 3.3794032125342963, "learning_rate": 6.163342830009497e-05, "loss": 0.5364, "step": 650 }, { "epoch": 0.31339031339031337, "grad_norm": 4.191975799476769, "learning_rate": 6.258309591642925e-05, "loss": 0.5208, "step": 660 }, { "epoch": 0.3181386514719848, "grad_norm": 3.85967469771651, "learning_rate": 6.353276353276355e-05, "loss": 0.4969, "step": 670 }, { "epoch": 0.32288698955365625, "grad_norm": 3.7614400963707024, "learning_rate": 6.448243114909783e-05, "loss": 0.4793, "step": 680 }, { "epoch": 0.32763532763532766, "grad_norm": 3.1186846516848337, "learning_rate": 6.54320987654321e-05, "loss": 0.4656, "step": 690 }, { "epoch": 0.33238366571699907, "grad_norm": 2.659191109701792, "learning_rate": 6.638176638176638e-05, "loss": 0.4521, "step": 700 }, { "epoch": 0.3371320037986705, "grad_norm": 3.4463157983016455, "learning_rate": 6.733143399810066e-05, "loss": 0.4321, "step": 710 }, { "epoch": 0.3418803418803419, "grad_norm": 3.5482062482332988, "learning_rate": 6.828110161443494e-05, "loss": 0.4161, "step": 720 }, { "epoch": 0.3466286799620133, "grad_norm": 3.3502617231342744, "learning_rate": 6.923076923076924e-05, "loss": 0.4054, "step": 730 }, { "epoch": 0.3513770180436847, "grad_norm": 3.195459439920149, "learning_rate": 7.018043684710352e-05, "loss": 0.383, "step": 740 }, { "epoch": 0.3561253561253561, "grad_norm": 3.4735859164121607, "learning_rate": 7.11301044634378e-05, "loss": 0.3744, "step": 750 }, { "epoch": 0.36087369420702753, "grad_norm": 3.175545355134774, "learning_rate": 7.207977207977208e-05, "loss": 0.3568, "step": 760 }, { "epoch": 0.36562203228869894, "grad_norm": 2.7379929250681068, "learning_rate": 7.302943969610636e-05, "loss": 0.3473, "step": 770 }, { "epoch": 0.37037037037037035, "grad_norm": 2.0806186299647216, "learning_rate": 7.397910731244065e-05, "loss": 0.3325, "step": 780 }, { "epoch": 0.37511870845204176, "grad_norm": 2.917840488213418, "learning_rate": 7.492877492877494e-05, "loss": 0.3174, "step": 790 }, { "epoch": 0.3798670465337132, "grad_norm": 2.5280776976703216, "learning_rate": 7.587844254510922e-05, "loss": 0.3083, "step": 800 }, { "epoch": 0.38461538461538464, "grad_norm": 4.449606742103996, "learning_rate": 7.68281101614435e-05, "loss": 0.308, "step": 810 }, { "epoch": 0.38936372269705605, "grad_norm": 2.5627464385183876, "learning_rate": 7.777777777777778e-05, "loss": 0.2965, "step": 820 }, { "epoch": 0.39411206077872746, "grad_norm": 1.8609497310412055, "learning_rate": 7.872744539411206e-05, "loss": 0.2708, "step": 830 }, { "epoch": 0.39886039886039887, "grad_norm": 2.102753950382446, "learning_rate": 7.967711301044634e-05, "loss": 0.2675, "step": 840 }, { "epoch": 0.4036087369420703, "grad_norm": 2.1996856846433395, "learning_rate": 8.062678062678063e-05, "loss": 0.2482, "step": 850 }, { "epoch": 0.4083570750237417, "grad_norm": 2.099118212528178, "learning_rate": 8.157644824311491e-05, "loss": 0.2368, "step": 860 }, { "epoch": 0.4131054131054131, "grad_norm": 1.6495997027020226, "learning_rate": 8.25261158594492e-05, "loss": 0.2251, "step": 870 }, { "epoch": 0.4178537511870845, "grad_norm": 1.6352958983011037, "learning_rate": 8.347578347578347e-05, "loss": 0.2154, "step": 880 }, { "epoch": 0.4226020892687559, "grad_norm": 1.2086967138392473, "learning_rate": 8.442545109211775e-05, "loss": 0.2097, "step": 890 }, { "epoch": 0.42735042735042733, "grad_norm": 2.295283958872144, "learning_rate": 8.537511870845205e-05, "loss": 0.2006, "step": 900 }, { "epoch": 0.43209876543209874, "grad_norm": 1.6079048977198918, "learning_rate": 8.632478632478634e-05, "loss": 0.1852, "step": 910 }, { "epoch": 0.4368471035137702, "grad_norm": 1.8132132209600655, "learning_rate": 8.727445394112062e-05, "loss": 0.177, "step": 920 }, { "epoch": 0.4415954415954416, "grad_norm": 1.6117371306924786, "learning_rate": 8.82241215574549e-05, "loss": 0.175, "step": 930 }, { "epoch": 0.44634377967711303, "grad_norm": 1.5896963948942289, "learning_rate": 8.917378917378918e-05, "loss": 0.1725, "step": 940 }, { "epoch": 0.45109211775878444, "grad_norm": 1.4416748631155214, "learning_rate": 9.012345679012346e-05, "loss": 0.1537, "step": 950 }, { "epoch": 0.45584045584045585, "grad_norm": 1.1506191039882334, "learning_rate": 9.107312440645774e-05, "loss": 0.1472, "step": 960 }, { "epoch": 0.46058879392212726, "grad_norm": 1.2792430834204105, "learning_rate": 9.202279202279203e-05, "loss": 0.1405, "step": 970 }, { "epoch": 0.46533713200379867, "grad_norm": 1.3038062291321308, "learning_rate": 9.297245963912631e-05, "loss": 0.1325, "step": 980 }, { "epoch": 0.4700854700854701, "grad_norm": 1.2877698203824843, "learning_rate": 9.392212725546059e-05, "loss": 0.1259, "step": 990 }, { "epoch": 0.4748338081671415, "grad_norm": 1.2411251108456811, "learning_rate": 9.487179487179487e-05, "loss": 0.1228, "step": 1000 }, { "epoch": 0.4795821462488129, "grad_norm": 1.2371657488544003, "learning_rate": 9.582146248812915e-05, "loss": 0.1157, "step": 1010 }, { "epoch": 0.4843304843304843, "grad_norm": 1.9824003783836297, "learning_rate": 9.677113010446344e-05, "loss": 0.115, "step": 1020 }, { "epoch": 0.4890788224121557, "grad_norm": 0.9270272735488306, "learning_rate": 9.772079772079774e-05, "loss": 0.1078, "step": 1030 }, { "epoch": 0.49382716049382713, "grad_norm": 1.69664894707242, "learning_rate": 9.867046533713202e-05, "loss": 0.1044, "step": 1040 }, { "epoch": 0.4985754985754986, "grad_norm": 0.7192930505558002, "learning_rate": 9.96201329534663e-05, "loss": 0.0972, "step": 1050 }, { "epoch": 0.50332383665717, "grad_norm": 1.0334470862657361, "learning_rate": 9.999990109907572e-05, "loss": 0.0933, "step": 1060 }, { "epoch": 0.5080721747388414, "grad_norm": 0.7970323005171996, "learning_rate": 9.999929670595522e-05, "loss": 0.0951, "step": 1070 }, { "epoch": 0.5128205128205128, "grad_norm": 2.103019694166396, "learning_rate": 9.99981428713058e-05, "loss": 0.093, "step": 1080 }, { "epoch": 0.5175688509021842, "grad_norm": 0.5875319206363789, "learning_rate": 9.999643960780693e-05, "loss": 0.0864, "step": 1090 }, { "epoch": 0.5223171889838556, "grad_norm": 0.5028113774010918, "learning_rate": 9.999418693417573e-05, "loss": 0.078, "step": 1100 }, { "epoch": 0.5270655270655271, "grad_norm": 0.5601155714961593, "learning_rate": 9.999138487516681e-05, "loss": 0.0719, "step": 1110 }, { "epoch": 0.5318138651471985, "grad_norm": 0.5733378588111542, "learning_rate": 9.998803346157195e-05, "loss": 0.071, "step": 1120 }, { "epoch": 0.5365622032288699, "grad_norm": 0.46057204449456696, "learning_rate": 9.99841327302198e-05, "loss": 0.0667, "step": 1130 }, { "epoch": 0.5413105413105413, "grad_norm": 0.40532486560347053, "learning_rate": 9.997968272397542e-05, "loss": 0.0646, "step": 1140 }, { "epoch": 0.5460588793922128, "grad_norm": 1.1210562619250999, "learning_rate": 9.997468349173985e-05, "loss": 0.0638, "step": 1150 }, { "epoch": 0.5508072174738842, "grad_norm": 0.6169955889548634, "learning_rate": 9.99691350884496e-05, "loss": 0.0632, "step": 1160 }, { "epoch": 0.5555555555555556, "grad_norm": 0.6578535651069357, "learning_rate": 9.996303757507602e-05, "loss": 0.0587, "step": 1170 }, { "epoch": 0.560303893637227, "grad_norm": 0.4171314533643107, "learning_rate": 9.995639101862453e-05, "loss": 0.0571, "step": 1180 }, { "epoch": 0.5650522317188984, "grad_norm": 0.38986655735794734, "learning_rate": 9.994919549213408e-05, "loss": 0.0547, "step": 1190 }, { "epoch": 0.5698005698005698, "grad_norm": 0.35679310646076673, "learning_rate": 9.99414510746762e-05, "loss": 0.0533, "step": 1200 }, { "epoch": 0.5745489078822412, "grad_norm": 0.9756351383699576, "learning_rate": 9.993315785135416e-05, "loss": 0.0519, "step": 1210 }, { "epoch": 0.5792972459639126, "grad_norm": 0.30453461215635685, "learning_rate": 9.99243159133021e-05, "loss": 0.0507, "step": 1220 }, { "epoch": 0.584045584045584, "grad_norm": 0.26889233181887895, "learning_rate": 9.991492535768392e-05, "loss": 0.0485, "step": 1230 }, { "epoch": 0.5887939221272555, "grad_norm": 0.44897606822377356, "learning_rate": 9.99049862876923e-05, "loss": 0.0479, "step": 1240 }, { "epoch": 0.5935422602089269, "grad_norm": 0.3323201059302271, "learning_rate": 9.989449881254754e-05, "loss": 0.0455, "step": 1250 }, { "epoch": 0.5982905982905983, "grad_norm": 0.3450397191457616, "learning_rate": 9.988346304749635e-05, "loss": 0.0459, "step": 1260 }, { "epoch": 0.6030389363722697, "grad_norm": 0.2915662683573587, "learning_rate": 9.987187911381059e-05, "loss": 0.0448, "step": 1270 }, { "epoch": 0.6077872744539411, "grad_norm": 0.27436376252679173, "learning_rate": 9.985974713878592e-05, "loss": 0.0432, "step": 1280 }, { "epoch": 0.6125356125356125, "grad_norm": 0.2971826802601137, "learning_rate": 9.984706725574044e-05, "loss": 0.0424, "step": 1290 }, { "epoch": 0.6172839506172839, "grad_norm": 0.30519567092051064, "learning_rate": 9.983383960401319e-05, "loss": 0.0409, "step": 1300 }, { "epoch": 0.6220322886989553, "grad_norm": 0.3418674827222748, "learning_rate": 9.982006432896266e-05, "loss": 0.0402, "step": 1310 }, { "epoch": 0.6267806267806267, "grad_norm": 0.29237867980087867, "learning_rate": 9.980574158196512e-05, "loss": 0.0396, "step": 1320 }, { "epoch": 0.6315289648622981, "grad_norm": 0.2601228902375644, "learning_rate": 9.979087152041304e-05, "loss": 0.0385, "step": 1330 }, { "epoch": 0.6362773029439696, "grad_norm": 0.3060864165416616, "learning_rate": 9.977545430771332e-05, "loss": 0.0374, "step": 1340 }, { "epoch": 0.6410256410256411, "grad_norm": 0.5433741092932173, "learning_rate": 9.975949011328542e-05, "loss": 0.0366, "step": 1350 }, { "epoch": 0.6457739791073125, "grad_norm": 0.2674649839145288, "learning_rate": 9.97429791125597e-05, "loss": 0.0356, "step": 1360 }, { "epoch": 0.6505223171889839, "grad_norm": 0.26496073204488824, "learning_rate": 9.972592148697528e-05, "loss": 0.0349, "step": 1370 }, { "epoch": 0.6552706552706553, "grad_norm": 0.28343162637710845, "learning_rate": 9.970831742397815e-05, "loss": 0.0337, "step": 1380 }, { "epoch": 0.6600189933523267, "grad_norm": 0.3066298313345385, "learning_rate": 9.969016711701908e-05, "loss": 0.033, "step": 1390 }, { "epoch": 0.6647673314339981, "grad_norm": 0.7302731671525841, "learning_rate": 9.967147076555157e-05, "loss": 0.0322, "step": 1400 }, { "epoch": 0.6695156695156695, "grad_norm": 0.7933940874417845, "learning_rate": 9.965222857502951e-05, "loss": 0.0395, "step": 1410 }, { "epoch": 0.674264007597341, "grad_norm": 0.2836389842951923, "learning_rate": 9.963244075690505e-05, "loss": 0.0359, "step": 1420 }, { "epoch": 0.6790123456790124, "grad_norm": 0.24878347396026249, "learning_rate": 9.961210752862626e-05, "loss": 0.0317, "step": 1430 }, { "epoch": 0.6837606837606838, "grad_norm": 0.22171700267192393, "learning_rate": 9.959122911363463e-05, "loss": 0.03, "step": 1440 }, { "epoch": 0.6885090218423552, "grad_norm": 0.24135787011018217, "learning_rate": 9.956980574136281e-05, "loss": 0.0291, "step": 1450 }, { "epoch": 0.6932573599240266, "grad_norm": 0.24244872011656793, "learning_rate": 9.954783764723188e-05, "loss": 0.0282, "step": 1460 }, { "epoch": 0.698005698005698, "grad_norm": 0.20881098829135122, "learning_rate": 9.952532507264891e-05, "loss": 0.0274, "step": 1470 }, { "epoch": 0.7027540360873694, "grad_norm": 0.1735175386102607, "learning_rate": 9.950226826500426e-05, "loss": 0.0263, "step": 1480 }, { "epoch": 0.7075023741690408, "grad_norm": 0.21392183402851034, "learning_rate": 9.947866747766882e-05, "loss": 0.0259, "step": 1490 }, { "epoch": 0.7122507122507122, "grad_norm": 0.23777008260247748, "learning_rate": 9.945452296999131e-05, "loss": 0.0255, "step": 1500 }, { "epoch": 0.7169990503323836, "grad_norm": 0.306026750469362, "learning_rate": 9.942983500729536e-05, "loss": 0.025, "step": 1510 }, { "epoch": 0.7217473884140551, "grad_norm": 0.2689307207882034, "learning_rate": 9.940460386087667e-05, "loss": 0.0245, "step": 1520 }, { "epoch": 0.7264957264957265, "grad_norm": 0.1975647423467133, "learning_rate": 9.937882980799986e-05, "loss": 0.0236, "step": 1530 }, { "epoch": 0.7312440645773979, "grad_norm": 0.20101737344080323, "learning_rate": 9.935251313189564e-05, "loss": 0.0231, "step": 1540 }, { "epoch": 0.7359924026590693, "grad_norm": 0.17695615800799033, "learning_rate": 9.93256541217576e-05, "loss": 0.0224, "step": 1550 }, { "epoch": 0.7407407407407407, "grad_norm": 0.4067301224979671, "learning_rate": 9.929825307273898e-05, "loss": 0.0224, "step": 1560 }, { "epoch": 0.7454890788224121, "grad_norm": 0.2862796956587721, "learning_rate": 9.927031028594951e-05, "loss": 0.0229, "step": 1570 }, { "epoch": 0.7502374169040835, "grad_norm": 0.17919163097786697, "learning_rate": 9.924182606845204e-05, "loss": 0.0224, "step": 1580 }, { "epoch": 0.7549857549857549, "grad_norm": 0.16203033039111173, "learning_rate": 9.921280073325922e-05, "loss": 0.0211, "step": 1590 }, { "epoch": 0.7597340930674265, "grad_norm": 0.15748963931516455, "learning_rate": 9.918323459933005e-05, "loss": 0.0202, "step": 1600 }, { "epoch": 0.7644824311490979, "grad_norm": 0.14768750521303628, "learning_rate": 9.915312799156632e-05, "loss": 0.0197, "step": 1610 }, { "epoch": 0.7692307692307693, "grad_norm": 0.1497846834623332, "learning_rate": 9.91224812408091e-05, "loss": 0.0194, "step": 1620 }, { "epoch": 0.7739791073124407, "grad_norm": 0.14462122268426925, "learning_rate": 9.909129468383506e-05, "loss": 0.0194, "step": 1630 }, { "epoch": 0.7787274453941121, "grad_norm": 0.1699059756635377, "learning_rate": 9.905956866335286e-05, "loss": 0.0187, "step": 1640 }, { "epoch": 0.7834757834757835, "grad_norm": 0.1384218259721326, "learning_rate": 9.90273035279992e-05, "loss": 0.0187, "step": 1650 }, { "epoch": 0.7882241215574549, "grad_norm": 0.16571495924373608, "learning_rate": 9.899449963233523e-05, "loss": 0.0183, "step": 1660 }, { "epoch": 0.7929724596391263, "grad_norm": 0.1364834427504214, "learning_rate": 9.896115733684244e-05, "loss": 0.0182, "step": 1670 }, { "epoch": 0.7977207977207977, "grad_norm": 0.1394284544093541, "learning_rate": 9.892727700791881e-05, "loss": 0.0174, "step": 1680 }, { "epoch": 0.8024691358024691, "grad_norm": 0.1419407740453267, "learning_rate": 9.889285901787482e-05, "loss": 0.0172, "step": 1690 }, { "epoch": 0.8072174738841406, "grad_norm": 0.1489778339389939, "learning_rate": 9.885790374492921e-05, "loss": 0.0173, "step": 1700 }, { "epoch": 0.811965811965812, "grad_norm": 0.14880857407194373, "learning_rate": 9.882241157320497e-05, "loss": 0.0171, "step": 1710 }, { "epoch": 0.8167141500474834, "grad_norm": 0.1312532199906158, "learning_rate": 9.878638289272507e-05, "loss": 0.0167, "step": 1720 }, { "epoch": 0.8214624881291548, "grad_norm": 0.18660180447513916, "learning_rate": 9.874981809940808e-05, "loss": 0.0164, "step": 1730 }, { "epoch": 0.8262108262108262, "grad_norm": 0.1504088979175503, "learning_rate": 9.871271759506404e-05, "loss": 0.0164, "step": 1740 }, { "epoch": 0.8309591642924976, "grad_norm": 0.16280752340641091, "learning_rate": 9.86750817873898e-05, "loss": 0.0159, "step": 1750 }, { "epoch": 0.835707502374169, "grad_norm": 0.1429005372176506, "learning_rate": 9.863691108996471e-05, "loss": 0.0161, "step": 1760 }, { "epoch": 0.8404558404558404, "grad_norm": 0.1414044554475113, "learning_rate": 9.859820592224599e-05, "loss": 0.0155, "step": 1770 }, { "epoch": 0.8452041785375118, "grad_norm": 0.12442609011687472, "learning_rate": 9.855896670956415e-05, "loss": 0.0157, "step": 1780 }, { "epoch": 0.8499525166191833, "grad_norm": 0.13787531332505618, "learning_rate": 9.851919388311832e-05, "loss": 0.0155, "step": 1790 }, { "epoch": 0.8547008547008547, "grad_norm": 0.14024759166486325, "learning_rate": 9.84788878799715e-05, "loss": 0.015, "step": 1800 }, { "epoch": 0.8594491927825261, "grad_norm": 0.14052815351143585, "learning_rate": 9.843804914304577e-05, "loss": 0.0152, "step": 1810 }, { "epoch": 0.8641975308641975, "grad_norm": 0.14285787402293643, "learning_rate": 9.839667812111739e-05, "loss": 0.0149, "step": 1820 }, { "epoch": 0.8689458689458689, "grad_norm": 0.14004431618365418, "learning_rate": 9.835477526881189e-05, "loss": 0.0149, "step": 1830 }, { "epoch": 0.8736942070275404, "grad_norm": 0.13456900134182037, "learning_rate": 9.831234104659913e-05, "loss": 0.0145, "step": 1840 }, { "epoch": 0.8784425451092118, "grad_norm": 0.13153333115946414, "learning_rate": 9.826937592078811e-05, "loss": 0.0144, "step": 1850 }, { "epoch": 0.8831908831908832, "grad_norm": 0.13111908038395156, "learning_rate": 9.822588036352201e-05, "loss": 0.0144, "step": 1860 }, { "epoch": 0.8879392212725546, "grad_norm": 0.13537448758577494, "learning_rate": 9.81818548527728e-05, "loss": 0.0141, "step": 1870 }, { "epoch": 0.8926875593542261, "grad_norm": 0.12336646433662755, "learning_rate": 9.81372998723362e-05, "loss": 0.0138, "step": 1880 }, { "epoch": 0.8974358974358975, "grad_norm": 0.14430917737826196, "learning_rate": 9.809221591182623e-05, "loss": 0.0139, "step": 1890 }, { "epoch": 0.9021842355175689, "grad_norm": 0.11734321855280797, "learning_rate": 9.804660346666984e-05, "loss": 0.0137, "step": 1900 }, { "epoch": 0.9069325735992403, "grad_norm": 0.11606764102802321, "learning_rate": 9.800046303810154e-05, "loss": 0.0137, "step": 1910 }, { "epoch": 0.9116809116809117, "grad_norm": 0.14000693052105856, "learning_rate": 9.795379513315779e-05, "loss": 0.0137, "step": 1920 }, { "epoch": 0.9164292497625831, "grad_norm": 0.11593240867247616, "learning_rate": 9.790660026467153e-05, "loss": 0.0132, "step": 1930 }, { "epoch": 0.9211775878442545, "grad_norm": 0.13357626140322432, "learning_rate": 9.785887895126644e-05, "loss": 0.0132, "step": 1940 }, { "epoch": 0.9259259259259259, "grad_norm": 0.13082008876183973, "learning_rate": 9.781063171735134e-05, "loss": 0.0132, "step": 1950 }, { "epoch": 0.9306742640075973, "grad_norm": 0.19902668639429527, "learning_rate": 9.776185909311437e-05, "loss": 0.0136, "step": 1960 }, { "epoch": 0.9354226020892688, "grad_norm": 0.14688628062037878, "learning_rate": 9.771256161451716e-05, "loss": 0.0132, "step": 1970 }, { "epoch": 0.9401709401709402, "grad_norm": 0.17311744484721647, "learning_rate": 9.766273982328899e-05, "loss": 0.013, "step": 1980 }, { "epoch": 0.9449192782526116, "grad_norm": 0.11972076423332602, "learning_rate": 9.761239426692077e-05, "loss": 0.0128, "step": 1990 }, { "epoch": 0.949667616334283, "grad_norm": 0.10904986633123297, "learning_rate": 9.756152549865907e-05, "loss": 0.0127, "step": 2000 }, { "epoch": 0.9544159544159544, "grad_norm": 0.12074328690288258, "learning_rate": 9.751013407750004e-05, "loss": 0.0127, "step": 2010 }, { "epoch": 0.9591642924976258, "grad_norm": 0.10314583447768316, "learning_rate": 9.745822056818324e-05, "loss": 0.0124, "step": 2020 }, { "epoch": 0.9639126305792972, "grad_norm": 0.11079053855550765, "learning_rate": 9.740578554118545e-05, "loss": 0.0126, "step": 2030 }, { "epoch": 0.9686609686609686, "grad_norm": 0.11039647449035304, "learning_rate": 9.735282957271438e-05, "loss": 0.0122, "step": 2040 }, { "epoch": 0.97340930674264, "grad_norm": 0.1074248823883716, "learning_rate": 9.729935324470243e-05, "loss": 0.0122, "step": 2050 }, { "epoch": 0.9781576448243114, "grad_norm": 0.11432167277350053, "learning_rate": 9.724535714480012e-05, "loss": 0.0122, "step": 2060 }, { "epoch": 0.9829059829059829, "grad_norm": 0.10327032601323838, "learning_rate": 9.719084186636984e-05, "loss": 0.0121, "step": 2070 }, { "epoch": 0.9876543209876543, "grad_norm": 0.10498334396065222, "learning_rate": 9.713580800847916e-05, "loss": 0.012, "step": 2080 }, { "epoch": 0.9924026590693258, "grad_norm": 0.12706630397842825, "learning_rate": 9.708025617589431e-05, "loss": 0.0119, "step": 2090 }, { "epoch": 0.9971509971509972, "grad_norm": 0.1386343949216736, "learning_rate": 9.702418697907357e-05, "loss": 0.0118, "step": 2100 }, { "epoch": 1.0018993352326686, "grad_norm": 0.17348868630481104, "learning_rate": 9.696760103416049e-05, "loss": 0.0118, "step": 2110 }, { "epoch": 1.00664767331434, "grad_norm": 0.10746029382832634, "learning_rate": 9.69104989629772e-05, "loss": 0.0115, "step": 2120 }, { "epoch": 1.0113960113960114, "grad_norm": 0.1098099997844564, "learning_rate": 9.685288139301749e-05, "loss": 0.0115, "step": 2130 }, { "epoch": 1.0161443494776827, "grad_norm": 0.10551311361504195, "learning_rate": 9.679474895744002e-05, "loss": 0.0114, "step": 2140 }, { "epoch": 1.0208926875593543, "grad_norm": 0.14702345368367736, "learning_rate": 9.673610229506123e-05, "loss": 0.0112, "step": 2150 }, { "epoch": 1.0256410256410255, "grad_norm": 0.10896703460263106, "learning_rate": 9.667694205034842e-05, "loss": 0.0113, "step": 2160 }, { "epoch": 1.030389363722697, "grad_norm": 0.12494149604296306, "learning_rate": 9.661726887341264e-05, "loss": 0.0111, "step": 2170 }, { "epoch": 1.0351377018043684, "grad_norm": 0.18134148447347836, "learning_rate": 9.655708342000158e-05, "loss": 0.0114, "step": 2180 }, { "epoch": 1.03988603988604, "grad_norm": 0.11150660791703594, "learning_rate": 9.649638635149225e-05, "loss": 0.0111, "step": 2190 }, { "epoch": 1.0446343779677112, "grad_norm": 0.16370516085173903, "learning_rate": 9.643517833488387e-05, "loss": 0.0112, "step": 2200 }, { "epoch": 1.0493827160493827, "grad_norm": 0.10999453847016634, "learning_rate": 9.637346004279038e-05, "loss": 0.011, "step": 2210 }, { "epoch": 1.0541310541310542, "grad_norm": 0.09968933250946793, "learning_rate": 9.631123215343323e-05, "loss": 0.011, "step": 2220 }, { "epoch": 1.0588793922127255, "grad_norm": 0.09940227976913124, "learning_rate": 9.624849535063373e-05, "loss": 0.0109, "step": 2230 }, { "epoch": 1.063627730294397, "grad_norm": 0.10701034026061397, "learning_rate": 9.61852503238057e-05, "loss": 0.0109, "step": 2240 }, { "epoch": 1.0683760683760684, "grad_norm": 0.09350833494425366, "learning_rate": 9.612149776794776e-05, "loss": 0.0109, "step": 2250 }, { "epoch": 1.0731244064577399, "grad_norm": 0.11271042172806738, "learning_rate": 9.605723838363586e-05, "loss": 0.0108, "step": 2260 }, { "epoch": 1.0778727445394112, "grad_norm": 0.10423498520674267, "learning_rate": 9.599247287701539e-05, "loss": 0.011, "step": 2270 }, { "epoch": 1.0826210826210827, "grad_norm": 0.10744159462037456, "learning_rate": 9.592720195979353e-05, "loss": 0.0109, "step": 2280 }, { "epoch": 1.087369420702754, "grad_norm": 0.12150767581378212, "learning_rate": 9.586142634923141e-05, "loss": 0.0108, "step": 2290 }, { "epoch": 1.0921177587844255, "grad_norm": 0.09934810940031444, "learning_rate": 9.579514676813627e-05, "loss": 0.0109, "step": 2300 }, { "epoch": 1.0968660968660968, "grad_norm": 0.18641934642746283, "learning_rate": 9.572836394485345e-05, "loss": 0.0129, "step": 2310 }, { "epoch": 1.1016144349477683, "grad_norm": 2.619600744042529, "learning_rate": 9.566107861325836e-05, "loss": 0.1392, "step": 2320 }, { "epoch": 1.1063627730294396, "grad_norm": 0.26072594767142904, "learning_rate": 9.559329151274854e-05, "loss": 0.0293, "step": 2330 }, { "epoch": 1.1111111111111112, "grad_norm": 0.15977598047524422, "learning_rate": 9.552500338823543e-05, "loss": 0.0153, "step": 2340 }, { "epoch": 1.1158594491927825, "grad_norm": 0.1255531165836577, "learning_rate": 9.54562149901362e-05, "loss": 0.0128, "step": 2350 }, { "epoch": 1.120607787274454, "grad_norm": 0.12323457866728675, "learning_rate": 9.538692707436554e-05, "loss": 0.0118, "step": 2360 }, { "epoch": 1.1253561253561253, "grad_norm": 0.09850210122582442, "learning_rate": 9.531714040232733e-05, "loss": 0.0116, "step": 2370 }, { "epoch": 1.1301044634377968, "grad_norm": 0.10055693463056198, "learning_rate": 9.524685574090627e-05, "loss": 0.0114, "step": 2380 }, { "epoch": 1.134852801519468, "grad_norm": 0.10061345862429888, "learning_rate": 9.517607386245948e-05, "loss": 0.0108, "step": 2390 }, { "epoch": 1.1396011396011396, "grad_norm": 0.09410721120726814, "learning_rate": 9.510479554480793e-05, "loss": 0.0107, "step": 2400 }, { "epoch": 1.144349477682811, "grad_norm": 0.09074333253540687, "learning_rate": 9.503302157122804e-05, "loss": 0.0104, "step": 2410 }, { "epoch": 1.1490978157644824, "grad_norm": 0.09417010633097983, "learning_rate": 9.496075273044293e-05, "loss": 0.0105, "step": 2420 }, { "epoch": 1.1538461538461537, "grad_norm": 0.08527272496006741, "learning_rate": 9.488798981661379e-05, "loss": 0.0104, "step": 2430 }, { "epoch": 1.1585944919278253, "grad_norm": 0.0745044104987499, "learning_rate": 9.481473362933122e-05, "loss": 0.0102, "step": 2440 }, { "epoch": 1.1633428300094968, "grad_norm": 0.08538110524754806, "learning_rate": 9.474098497360637e-05, "loss": 0.0101, "step": 2450 }, { "epoch": 1.168091168091168, "grad_norm": 0.08771994737551454, "learning_rate": 9.466674465986211e-05, "loss": 0.0103, "step": 2460 }, { "epoch": 1.1728395061728394, "grad_norm": 0.10081585564691586, "learning_rate": 9.459201350392418e-05, "loss": 0.0103, "step": 2470 }, { "epoch": 1.177587844254511, "grad_norm": 0.08629118740719248, "learning_rate": 9.451679232701212e-05, "loss": 0.0101, "step": 2480 }, { "epoch": 1.1823361823361824, "grad_norm": 0.07776803009416808, "learning_rate": 9.444108195573031e-05, "loss": 0.01, "step": 2490 }, { "epoch": 1.1870845204178537, "grad_norm": 0.09455703315561588, "learning_rate": 9.436488322205892e-05, "loss": 0.0103, "step": 2500 }, { "epoch": 1.1918328584995252, "grad_norm": 0.1076036848212035, "learning_rate": 9.42881969633447e-05, "loss": 0.0102, "step": 2510 }, { "epoch": 1.1965811965811965, "grad_norm": 0.09744309522609514, "learning_rate": 9.42110240222918e-05, "loss": 0.0099, "step": 2520 }, { "epoch": 1.201329534662868, "grad_norm": 0.08682352879189964, "learning_rate": 9.413336524695253e-05, "loss": 0.0101, "step": 2530 }, { "epoch": 1.2060778727445394, "grad_norm": 0.08237507848311064, "learning_rate": 9.405522149071801e-05, "loss": 0.0099, "step": 2540 }, { "epoch": 1.210826210826211, "grad_norm": 0.08700855379798583, "learning_rate": 9.397659361230883e-05, "loss": 0.0097, "step": 2550 }, { "epoch": 1.2155745489078822, "grad_norm": 0.08866861899387227, "learning_rate": 9.389748247576557e-05, "loss": 0.0097, "step": 2560 }, { "epoch": 1.2203228869895537, "grad_norm": 0.08986965854900687, "learning_rate": 9.381788895043933e-05, "loss": 0.01, "step": 2570 }, { "epoch": 1.225071225071225, "grad_norm": 0.07866337360055012, "learning_rate": 9.373781391098218e-05, "loss": 0.0098, "step": 2580 }, { "epoch": 1.2298195631528965, "grad_norm": 0.0724649262291167, "learning_rate": 9.365725823733753e-05, "loss": 0.0096, "step": 2590 }, { "epoch": 1.2345679012345678, "grad_norm": 0.08012257907966706, "learning_rate": 9.35762228147305e-05, "loss": 0.0097, "step": 2600 }, { "epoch": 1.2393162393162394, "grad_norm": 0.0728449403624035, "learning_rate": 9.349470853365815e-05, "loss": 0.0095, "step": 2610 }, { "epoch": 1.2440645773979107, "grad_norm": 0.08436770269446454, "learning_rate": 9.341271628987968e-05, "loss": 0.0097, "step": 2620 }, { "epoch": 1.2488129154795822, "grad_norm": 0.08804859686962187, "learning_rate": 9.333024698440667e-05, "loss": 0.0095, "step": 2630 }, { "epoch": 1.2535612535612537, "grad_norm": 0.07983530496221104, "learning_rate": 9.324730152349306e-05, "loss": 0.0095, "step": 2640 }, { "epoch": 1.258309591642925, "grad_norm": 0.06877839009766883, "learning_rate": 9.31638808186253e-05, "loss": 0.0095, "step": 2650 }, { "epoch": 1.2630579297245963, "grad_norm": 0.07520592231353386, "learning_rate": 9.307998578651228e-05, "loss": 0.0093, "step": 2660 }, { "epoch": 1.2678062678062678, "grad_norm": 0.07738504320294105, "learning_rate": 9.299561734907528e-05, "loss": 0.0093, "step": 2670 }, { "epoch": 1.2725546058879393, "grad_norm": 0.07419782478261074, "learning_rate": 9.291077643343776e-05, "loss": 0.0095, "step": 2680 }, { "epoch": 1.2773029439696106, "grad_norm": 0.06819785697023671, "learning_rate": 9.282546397191531e-05, "loss": 0.0094, "step": 2690 }, { "epoch": 1.282051282051282, "grad_norm": 0.07744020238479, "learning_rate": 9.273968090200533e-05, "loss": 0.0094, "step": 2700 }, { "epoch": 1.2867996201329535, "grad_norm": 0.07131447608372647, "learning_rate": 9.265342816637666e-05, "loss": 0.0093, "step": 2710 }, { "epoch": 1.291547958214625, "grad_norm": 0.06987449196589228, "learning_rate": 9.256670671285934e-05, "loss": 0.0091, "step": 2720 }, { "epoch": 1.2962962962962963, "grad_norm": 0.07956054671934792, "learning_rate": 9.247951749443413e-05, "loss": 0.0094, "step": 2730 }, { "epoch": 1.3010446343779676, "grad_norm": 0.07788375514648964, "learning_rate": 9.239186146922208e-05, "loss": 0.0093, "step": 2740 }, { "epoch": 1.305792972459639, "grad_norm": 0.0680950545496221, "learning_rate": 9.230373960047394e-05, "loss": 0.0092, "step": 2750 }, { "epoch": 1.3105413105413106, "grad_norm": 0.062302712796328624, "learning_rate": 9.22151528565596e-05, "loss": 0.0092, "step": 2760 }, { "epoch": 1.315289648622982, "grad_norm": 0.06891713045556644, "learning_rate": 9.212610221095748e-05, "loss": 0.0091, "step": 2770 }, { "epoch": 1.3200379867046534, "grad_norm": 0.06870146340482543, "learning_rate": 9.203658864224382e-05, "loss": 0.009, "step": 2780 }, { "epoch": 1.3247863247863247, "grad_norm": 0.06610061431786453, "learning_rate": 9.194661313408187e-05, "loss": 0.009, "step": 2790 }, { "epoch": 1.3295346628679963, "grad_norm": 0.07003109059654314, "learning_rate": 9.185617667521117e-05, "loss": 0.009, "step": 2800 }, { "epoch": 1.3342830009496676, "grad_norm": 0.09012523179481526, "learning_rate": 9.176528025943663e-05, "loss": 0.0093, "step": 2810 }, { "epoch": 1.339031339031339, "grad_norm": 0.07593213526212414, "learning_rate": 9.167392488561758e-05, "loss": 0.009, "step": 2820 }, { "epoch": 1.3437796771130104, "grad_norm": 0.07087160824813951, "learning_rate": 9.158211155765692e-05, "loss": 0.009, "step": 2830 }, { "epoch": 1.348528015194682, "grad_norm": 0.0741500373572918, "learning_rate": 9.14898412844899e-05, "loss": 0.009, "step": 2840 }, { "epoch": 1.3532763532763532, "grad_norm": 0.07471272709713543, "learning_rate": 9.13971150800732e-05, "loss": 0.0089, "step": 2850 }, { "epoch": 1.3580246913580247, "grad_norm": 0.12863323060835957, "learning_rate": 9.130393396337368e-05, "loss": 0.0091, "step": 2860 }, { "epoch": 1.362773029439696, "grad_norm": 0.09421023991900089, "learning_rate": 9.121029895835731e-05, "loss": 0.0092, "step": 2870 }, { "epoch": 1.3675213675213675, "grad_norm": 0.08288888706606813, "learning_rate": 9.111621109397769e-05, "loss": 0.0092, "step": 2880 }, { "epoch": 1.3722697056030388, "grad_norm": 0.06898757070782531, "learning_rate": 9.102167140416503e-05, "loss": 0.0091, "step": 2890 }, { "epoch": 1.3770180436847104, "grad_norm": 0.0654593731547835, "learning_rate": 9.092668092781454e-05, "loss": 0.0088, "step": 2900 }, { "epoch": 1.381766381766382, "grad_norm": 0.08022874035281295, "learning_rate": 9.083124070877516e-05, "loss": 0.0089, "step": 2910 }, { "epoch": 1.3865147198480532, "grad_norm": 0.07995153193335466, "learning_rate": 9.073535179583806e-05, "loss": 0.009, "step": 2920 }, { "epoch": 1.3912630579297245, "grad_norm": 0.08034340523525106, "learning_rate": 9.063901524272505e-05, "loss": 0.0089, "step": 2930 }, { "epoch": 1.396011396011396, "grad_norm": 0.07981592115944557, "learning_rate": 9.054223210807709e-05, "loss": 0.009, "step": 2940 }, { "epoch": 1.4007597340930675, "grad_norm": 0.07431367065596012, "learning_rate": 9.044500345544262e-05, "loss": 0.0088, "step": 2950 }, { "epoch": 1.4055080721747388, "grad_norm": 0.0882148739940019, "learning_rate": 9.034733035326584e-05, "loss": 0.0089, "step": 2960 }, { "epoch": 1.4102564102564101, "grad_norm": 0.062046125541639595, "learning_rate": 9.024921387487504e-05, "loss": 0.0089, "step": 2970 }, { "epoch": 1.4150047483380817, "grad_norm": 0.08387088640289536, "learning_rate": 9.015065509847075e-05, "loss": 0.0089, "step": 2980 }, { "epoch": 1.4197530864197532, "grad_norm": 0.06309895420956538, "learning_rate": 9.005165510711387e-05, "loss": 0.0088, "step": 2990 }, { "epoch": 1.4245014245014245, "grad_norm": 0.07285489878017845, "learning_rate": 8.995221498871387e-05, "loss": 0.0088, "step": 3000 }, { "epoch": 1.429249762583096, "grad_norm": 0.09206426204348243, "learning_rate": 8.98523358360167e-05, "loss": 0.0089, "step": 3010 }, { "epoch": 1.4339981006647673, "grad_norm": 0.07808601164973648, "learning_rate": 8.975201874659292e-05, "loss": 0.0089, "step": 3020 }, { "epoch": 1.4387464387464388, "grad_norm": 0.08547947351685176, "learning_rate": 8.965126482282549e-05, "loss": 0.0088, "step": 3030 }, { "epoch": 1.4434947768281101, "grad_norm": 0.06304686774614106, "learning_rate": 8.955007517189784e-05, "loss": 0.0086, "step": 3040 }, { "epoch": 1.4482431149097816, "grad_norm": 0.06690013381006199, "learning_rate": 8.944845090578149e-05, "loss": 0.0087, "step": 3050 }, { "epoch": 1.452991452991453, "grad_norm": 0.08260455005139142, "learning_rate": 8.934639314122401e-05, "loss": 0.0088, "step": 3060 }, { "epoch": 1.4577397910731245, "grad_norm": 0.06154823945097421, "learning_rate": 8.924390299973666e-05, "loss": 0.0087, "step": 3070 }, { "epoch": 1.4624881291547958, "grad_norm": 0.06929903681806748, "learning_rate": 8.91409816075821e-05, "loss": 0.0087, "step": 3080 }, { "epoch": 1.4672364672364673, "grad_norm": 0.20849055793591886, "learning_rate": 8.903763009576195e-05, "loss": 0.0091, "step": 3090 }, { "epoch": 1.4719848053181386, "grad_norm": 0.1784920261902527, "learning_rate": 8.893384960000445e-05, "loss": 0.0093, "step": 3100 }, { "epoch": 1.47673314339981, "grad_norm": 0.08143550680405819, "learning_rate": 8.882964126075192e-05, "loss": 0.0089, "step": 3110 }, { "epoch": 1.4814814814814814, "grad_norm": 0.09029117734458694, "learning_rate": 8.872500622314829e-05, "loss": 0.0088, "step": 3120 }, { "epoch": 1.486229819563153, "grad_norm": 0.081932882811764, "learning_rate": 8.861994563702638e-05, "loss": 0.0087, "step": 3130 }, { "epoch": 1.4909781576448244, "grad_norm": 0.08505267084725678, "learning_rate": 8.851446065689544e-05, "loss": 0.0086, "step": 3140 }, { "epoch": 1.4957264957264957, "grad_norm": 0.07153441399994394, "learning_rate": 8.840855244192836e-05, "loss": 0.0085, "step": 3150 }, { "epoch": 1.500474833808167, "grad_norm": 0.09944219276921275, "learning_rate": 8.83022221559489e-05, "loss": 0.0088, "step": 3160 }, { "epoch": 1.5052231718898386, "grad_norm": 0.07649587045866293, "learning_rate": 8.819547096741904e-05, "loss": 0.0087, "step": 3170 }, { "epoch": 1.50997150997151, "grad_norm": 0.09283885484044746, "learning_rate": 8.8088300049426e-05, "loss": 0.0087, "step": 3180 }, { "epoch": 1.5147198480531814, "grad_norm": 0.06957062245943367, "learning_rate": 8.798071057966935e-05, "loss": 0.0087, "step": 3190 }, { "epoch": 1.5194681861348527, "grad_norm": 0.06397948473632416, "learning_rate": 8.787270374044824e-05, "loss": 0.0086, "step": 3200 }, { "epoch": 1.5242165242165242, "grad_norm": 0.06568594090780461, "learning_rate": 8.776428071864819e-05, "loss": 0.0086, "step": 3210 }, { "epoch": 1.5289648622981957, "grad_norm": 0.06255555459808987, "learning_rate": 8.76554427057282e-05, "loss": 0.0086, "step": 3220 }, { "epoch": 1.533713200379867, "grad_norm": 0.07583026108335658, "learning_rate": 8.754619089770755e-05, "loss": 0.0086, "step": 3230 }, { "epoch": 1.5384615384615383, "grad_norm": 0.06918772061270947, "learning_rate": 8.743652649515275e-05, "loss": 0.0085, "step": 3240 }, { "epoch": 1.5432098765432098, "grad_norm": 0.08032530757532322, "learning_rate": 8.73264507031643e-05, "loss": 0.0085, "step": 3250 }, { "epoch": 1.5479582146248814, "grad_norm": 0.0608584131456061, "learning_rate": 8.721596473136342e-05, "loss": 0.0085, "step": 3260 }, { "epoch": 1.5527065527065527, "grad_norm": 0.06934333742791383, "learning_rate": 8.710506979387882e-05, "loss": 0.0088, "step": 3270 }, { "epoch": 1.557454890788224, "grad_norm": 0.06856843730907558, "learning_rate": 8.699376710933336e-05, "loss": 0.0086, "step": 3280 }, { "epoch": 1.5622032288698955, "grad_norm": 0.07613868983212806, "learning_rate": 8.688205790083053e-05, "loss": 0.0086, "step": 3290 }, { "epoch": 1.566951566951567, "grad_norm": 0.06850348703912577, "learning_rate": 8.676994339594118e-05, "loss": 0.0085, "step": 3300 }, { "epoch": 1.5716999050332383, "grad_norm": 0.056149516186035114, "learning_rate": 8.665742482668996e-05, "loss": 0.0083, "step": 3310 }, { "epoch": 1.5764482431149098, "grad_norm": 0.07910024549589795, "learning_rate": 8.654450342954175e-05, "loss": 0.0084, "step": 3320 }, { "epoch": 1.5811965811965814, "grad_norm": 0.06858204680711835, "learning_rate": 8.643118044538808e-05, "loss": 0.0084, "step": 3330 }, { "epoch": 1.5859449192782527, "grad_norm": 0.06876634599528418, "learning_rate": 8.631745711953357e-05, "loss": 0.0083, "step": 3340 }, { "epoch": 1.590693257359924, "grad_norm": 0.07101332966376236, "learning_rate": 8.620333470168214e-05, "loss": 0.0084, "step": 3350 }, { "epoch": 1.5954415954415955, "grad_norm": 0.0741047529646423, "learning_rate": 8.608881444592337e-05, "loss": 0.0084, "step": 3360 }, { "epoch": 1.600189933523267, "grad_norm": 0.06873949299023203, "learning_rate": 8.597389761071862e-05, "loss": 0.0083, "step": 3370 }, { "epoch": 1.6049382716049383, "grad_norm": 0.06882345991099557, "learning_rate": 8.58585854588873e-05, "loss": 0.0083, "step": 3380 }, { "epoch": 1.6096866096866096, "grad_norm": 0.05740311981903638, "learning_rate": 8.574287925759294e-05, "loss": 0.0084, "step": 3390 }, { "epoch": 1.6144349477682811, "grad_norm": 0.06425469647219727, "learning_rate": 8.562678027832925e-05, "loss": 0.0084, "step": 3400 }, { "epoch": 1.6191832858499526, "grad_norm": 0.06513521309053781, "learning_rate": 8.551028979690621e-05, "loss": 0.0083, "step": 3410 }, { "epoch": 1.623931623931624, "grad_norm": 0.06451656679873198, "learning_rate": 8.539340909343597e-05, "loss": 0.0082, "step": 3420 }, { "epoch": 1.6286799620132952, "grad_norm": 0.060287906157322685, "learning_rate": 8.527613945231885e-05, "loss": 0.0083, "step": 3430 }, { "epoch": 1.6334283000949668, "grad_norm": 0.05435200184209839, "learning_rate": 8.51584821622292e-05, "loss": 0.008, "step": 3440 }, { "epoch": 1.6381766381766383, "grad_norm": 0.19981921156643304, "learning_rate": 8.504043851610122e-05, "loss": 0.0081, "step": 3450 }, { "epoch": 1.6429249762583096, "grad_norm": 2.8181059949404808, "learning_rate": 8.492200981111482e-05, "loss": 0.0652, "step": 3460 }, { "epoch": 1.6476733143399809, "grad_norm": 0.31988948204458406, "learning_rate": 8.480319734868124e-05, "loss": 0.047, "step": 3470 }, { "epoch": 1.6524216524216524, "grad_norm": 0.1460055769438615, "learning_rate": 8.468400243442888e-05, "loss": 0.0146, "step": 3480 }, { "epoch": 1.657169990503324, "grad_norm": 0.09698949806477933, "learning_rate": 8.456442637818889e-05, "loss": 0.0107, "step": 3490 }, { "epoch": 1.6619183285849952, "grad_norm": 0.08038765547861314, "learning_rate": 8.444447049398076e-05, "loss": 0.0097, "step": 3500 }, { "epoch": 1.6666666666666665, "grad_norm": 0.09377988125174774, "learning_rate": 8.432413609999794e-05, "loss": 0.0091, "step": 3510 }, { "epoch": 1.671415004748338, "grad_norm": 0.0769530374902733, "learning_rate": 8.420342451859329e-05, "loss": 0.0088, "step": 3520 }, { "epoch": 1.6761633428300096, "grad_norm": 0.06213114923488264, "learning_rate": 8.408233707626458e-05, "loss": 0.0086, "step": 3530 }, { "epoch": 1.6809116809116809, "grad_norm": 0.07651546836679228, "learning_rate": 8.39608751036399e-05, "loss": 0.0087, "step": 3540 }, { "epoch": 1.6856600189933522, "grad_norm": 0.05766706147018749, "learning_rate": 8.383903993546311e-05, "loss": 0.0086, "step": 3550 }, { "epoch": 1.690408357075024, "grad_norm": 0.06186749663592901, "learning_rate": 8.371683291057902e-05, "loss": 0.0083, "step": 3560 }, { "epoch": 1.6951566951566952, "grad_norm": 0.06452094370184336, "learning_rate": 8.359425537191883e-05, "loss": 0.0084, "step": 3570 }, { "epoch": 1.6999050332383665, "grad_norm": 0.06931867130446329, "learning_rate": 8.347130866648528e-05, "loss": 0.0083, "step": 3580 }, { "epoch": 1.704653371320038, "grad_norm": 0.05728364848326649, "learning_rate": 8.334799414533791e-05, "loss": 0.0082, "step": 3590 }, { "epoch": 1.7094017094017095, "grad_norm": 0.05897853836151738, "learning_rate": 8.322431316357815e-05, "loss": 0.0081, "step": 3600 } ], "logging_steps": 10, "max_steps": 10530, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 1800, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 337040490627072.0, "train_batch_size": 128, "trial_name": null, "trial_params": null }