{ "best_metric": null, "best_model_checkpoint": null, "epoch": 2.9839612085042893, "eval_steps": 2000, "global_step": 210000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.007104669544057832, "grad_norm": 9.946511268615723, "learning_rate": 1.1604224885141856e-06, "loss": 5.8176, "step": 500 }, { "epoch": 0.014209339088115664, "grad_norm": 14.786874771118164, "learning_rate": 2.3421588594704685e-06, "loss": 5.7587, "step": 1000 }, { "epoch": 0.021314008632173496, "grad_norm": 16.701108932495117, "learning_rate": 3.5262634395869846e-06, "loss": 5.532, "step": 1500 }, { "epoch": 0.028418678176231327, "grad_norm": 23.213130950927734, "learning_rate": 4.710368019703501e-06, "loss": 5.1119, "step": 2000 }, { "epoch": 0.028418678176231327, "eval_runtime": 930.4679, "eval_samples_per_second": 151.27, "eval_steps_per_second": 37.818, "step": 2000 }, { "epoch": 0.03552334772028916, "grad_norm": 23.64911651611328, "learning_rate": 5.894472599820016e-06, "loss": 4.4046, "step": 2500 }, { "epoch": 0.04262801726434699, "grad_norm": 27.94416046142578, "learning_rate": 7.078577179936533e-06, "loss": 4.1391, "step": 3000 }, { "epoch": 0.04973268680840483, "grad_norm": 25.3529109954834, "learning_rate": 8.260313550892814e-06, "loss": 4.0939, "step": 3500 }, { "epoch": 0.056837356352462655, "grad_norm": 36.91496658325195, "learning_rate": 9.444418131009332e-06, "loss": 3.9517, "step": 4000 }, { "epoch": 0.056837356352462655, "eval_runtime": 934.9504, "eval_samples_per_second": 150.545, "eval_steps_per_second": 37.636, "step": 4000 }, { "epoch": 0.06394202589652048, "grad_norm": 19.763195037841797, "learning_rate": 1.0628522711125847e-05, "loss": 3.9683, "step": 4500 }, { "epoch": 0.07104669544057832, "grad_norm": 22.828413009643555, "learning_rate": 1.1812627291242363e-05, "loss": 3.9331, "step": 5000 }, { "epoch": 0.07815136498463615, "grad_norm": 30.304622650146484, "learning_rate": 1.299673187135888e-05, "loss": 3.9017, "step": 5500 }, { "epoch": 0.08525603452869399, "grad_norm": 34.04104995727539, "learning_rate": 1.4180836451475396e-05, "loss": 3.8953, "step": 6000 }, { "epoch": 0.08525603452869399, "eval_runtime": 932.5246, "eval_samples_per_second": 150.937, "eval_steps_per_second": 37.734, "step": 6000 }, { "epoch": 0.09236070407275182, "grad_norm": 34.60411071777344, "learning_rate": 1.536494103159191e-05, "loss": 3.8456, "step": 6500 }, { "epoch": 0.09946537361680965, "grad_norm": 28.492341995239258, "learning_rate": 1.6549045611708428e-05, "loss": 3.8317, "step": 7000 }, { "epoch": 0.10657004316086748, "grad_norm": 25.529743194580078, "learning_rate": 1.7733150191824944e-05, "loss": 3.7857, "step": 7500 }, { "epoch": 0.11367471270492531, "grad_norm": 25.264846801757812, "learning_rate": 1.8914886562781225e-05, "loss": 3.7997, "step": 8000 }, { "epoch": 0.11367471270492531, "eval_runtime": 900.9582, "eval_samples_per_second": 156.225, "eval_steps_per_second": 39.056, "step": 8000 }, { "epoch": 0.12077938224898314, "grad_norm": 21.667879104614258, "learning_rate": 2.0098991142897742e-05, "loss": 3.8116, "step": 8500 }, { "epoch": 0.12788405179304096, "grad_norm": 24.545040130615234, "learning_rate": 2.128309572301426e-05, "loss": 3.8319, "step": 9000 }, { "epoch": 0.1349887213370988, "grad_norm": 23.32590675354004, "learning_rate": 2.246720030313077e-05, "loss": 3.785, "step": 9500 }, { "epoch": 0.14209339088115663, "grad_norm": 25.292354583740234, "learning_rate": 2.3651304883247288e-05, "loss": 3.7558, "step": 10000 }, { "epoch": 0.14209339088115663, "eval_runtime": 948.3831, "eval_samples_per_second": 148.413, "eval_steps_per_second": 37.103, "step": 10000 }, { "epoch": 0.14919806042521447, "grad_norm": 32.933631896972656, "learning_rate": 2.4833041254203573e-05, "loss": 3.6906, "step": 10500 }, { "epoch": 0.1563027299692723, "grad_norm": 32.69184494018555, "learning_rate": 2.601714583432009e-05, "loss": 3.8022, "step": 11000 }, { "epoch": 0.16340739951333014, "grad_norm": 22.242897033691406, "learning_rate": 2.7201250414436602e-05, "loss": 3.7729, "step": 11500 }, { "epoch": 0.17051206905738797, "grad_norm": 22.738082885742188, "learning_rate": 2.8385354994553122e-05, "loss": 3.7329, "step": 12000 }, { "epoch": 0.17051206905738797, "eval_runtime": 921.4048, "eval_samples_per_second": 152.758, "eval_steps_per_second": 38.19, "step": 12000 }, { "epoch": 0.1776167386014458, "grad_norm": 20.65458869934082, "learning_rate": 2.956945957466964e-05, "loss": 3.7221, "step": 12500 }, { "epoch": 0.18472140814550364, "grad_norm": 23.19024658203125, "learning_rate": 3.075119594562592e-05, "loss": 3.7592, "step": 13000 }, { "epoch": 0.19182607768956147, "grad_norm": 23.87779426574707, "learning_rate": 3.193530052574243e-05, "loss": 3.7483, "step": 13500 }, { "epoch": 0.1989307472336193, "grad_norm": 17.06186866760254, "learning_rate": 3.311703689669872e-05, "loss": 3.7152, "step": 14000 }, { "epoch": 0.1989307472336193, "eval_runtime": 932.5402, "eval_samples_per_second": 150.934, "eval_steps_per_second": 37.733, "step": 14000 }, { "epoch": 0.20603541677767712, "grad_norm": 28.80001449584961, "learning_rate": 3.430114147681524e-05, "loss": 3.6807, "step": 14500 }, { "epoch": 0.21314008632173495, "grad_norm": 23.454565048217773, "learning_rate": 3.548524605693175e-05, "loss": 3.7147, "step": 15000 }, { "epoch": 0.22024475586579279, "grad_norm": 23.109294891357422, "learning_rate": 3.6669350637048263e-05, "loss": 3.7469, "step": 15500 }, { "epoch": 0.22734942540985062, "grad_norm": 25.332721710205078, "learning_rate": 3.7853455217164777e-05, "loss": 3.701, "step": 16000 }, { "epoch": 0.22734942540985062, "eval_runtime": 926.1025, "eval_samples_per_second": 151.983, "eval_steps_per_second": 37.996, "step": 16000 }, { "epoch": 0.23445409495390845, "grad_norm": 18.929309844970703, "learning_rate": 3.9037559797281296e-05, "loss": 3.6502, "step": 16500 }, { "epoch": 0.2415587644979663, "grad_norm": 23.181020736694336, "learning_rate": 4.0221664377397816e-05, "loss": 3.7512, "step": 17000 }, { "epoch": 0.24866343404202412, "grad_norm": 23.910934448242188, "learning_rate": 4.140576895751433e-05, "loss": 3.6642, "step": 17500 }, { "epoch": 0.25576810358608193, "grad_norm": 21.21150779724121, "learning_rate": 4.258987353763085e-05, "loss": 3.6602, "step": 18000 }, { "epoch": 0.25576810358608193, "eval_runtime": 918.3944, "eval_samples_per_second": 153.259, "eval_steps_per_second": 38.315, "step": 18000 }, { "epoch": 0.2628727731301398, "grad_norm": 37.84225845336914, "learning_rate": 4.377397811774736e-05, "loss": 3.7027, "step": 18500 }, { "epoch": 0.2699774426741976, "grad_norm": 25.410938262939453, "learning_rate": 4.495808269786388e-05, "loss": 3.6725, "step": 19000 }, { "epoch": 0.27708211221825546, "grad_norm": 28.328292846679688, "learning_rate": 4.6142187277980395e-05, "loss": 3.6763, "step": 19500 }, { "epoch": 0.28418678176231327, "grad_norm": 23.013893127441406, "learning_rate": 4.732629185809691e-05, "loss": 3.6533, "step": 20000 }, { "epoch": 0.28418678176231327, "eval_runtime": 937.5319, "eval_samples_per_second": 150.13, "eval_steps_per_second": 37.533, "step": 20000 }, { "epoch": 0.29129145130637113, "grad_norm": 24.00814437866211, "learning_rate": 4.850802822905319e-05, "loss": 3.6695, "step": 20500 }, { "epoch": 0.29839612085042894, "grad_norm": 19.818140029907227, "learning_rate": 4.9692132809169706e-05, "loss": 3.6463, "step": 21000 }, { "epoch": 0.3055007903944868, "grad_norm": 16.67455291748047, "learning_rate": 4.9902639265321164e-05, "loss": 3.685, "step": 21500 }, { "epoch": 0.3126054599385446, "grad_norm": 53.42991638183594, "learning_rate": 4.977107070494435e-05, "loss": 3.695, "step": 22000 }, { "epoch": 0.3126054599385446, "eval_runtime": 943.5405, "eval_samples_per_second": 149.174, "eval_steps_per_second": 37.294, "step": 22000 }, { "epoch": 0.31971012948260247, "grad_norm": 23.91082000732422, "learning_rate": 4.963976528168829e-05, "loss": 3.6918, "step": 22500 }, { "epoch": 0.3268147990266603, "grad_norm": 24.224802017211914, "learning_rate": 4.9508196721311476e-05, "loss": 3.6241, "step": 23000 }, { "epoch": 0.3339194685707181, "grad_norm": 20.906675338745117, "learning_rate": 4.937662816093467e-05, "loss": 3.6723, "step": 23500 }, { "epoch": 0.34102413811477594, "grad_norm": 21.082576751708984, "learning_rate": 4.9245059600557854e-05, "loss": 3.623, "step": 24000 }, { "epoch": 0.34102413811477594, "eval_runtime": 932.7328, "eval_samples_per_second": 150.903, "eval_steps_per_second": 37.726, "step": 24000 }, { "epoch": 0.34812880765883375, "grad_norm": 21.17667579650879, "learning_rate": 4.911349104018104e-05, "loss": 3.6673, "step": 24500 }, { "epoch": 0.3552334772028916, "grad_norm": 16.182199478149414, "learning_rate": 4.898218561692498e-05, "loss": 3.6377, "step": 25000 }, { "epoch": 0.3623381467469494, "grad_norm": 18.022315979003906, "learning_rate": 4.885061705654817e-05, "loss": 3.6189, "step": 25500 }, { "epoch": 0.3694428162910073, "grad_norm": 21.209049224853516, "learning_rate": 4.871904849617135e-05, "loss": 3.6586, "step": 26000 }, { "epoch": 0.3694428162910073, "eval_runtime": 937.8314, "eval_samples_per_second": 150.082, "eval_steps_per_second": 37.521, "step": 26000 }, { "epoch": 0.3765474858350651, "grad_norm": 15.896419525146484, "learning_rate": 4.8587479935794544e-05, "loss": 3.6251, "step": 26500 }, { "epoch": 0.38365215537912295, "grad_norm": 18.70542335510254, "learning_rate": 4.8456174512538485e-05, "loss": 3.6287, "step": 27000 }, { "epoch": 0.39075682492318076, "grad_norm": 16.86925506591797, "learning_rate": 4.832460595216167e-05, "loss": 3.6281, "step": 27500 }, { "epoch": 0.3978614944672386, "grad_norm": 22.52922821044922, "learning_rate": 4.819303739178486e-05, "loss": 3.6099, "step": 28000 }, { "epoch": 0.3978614944672386, "eval_runtime": 917.9919, "eval_samples_per_second": 153.326, "eval_steps_per_second": 38.331, "step": 28000 }, { "epoch": 0.4049661640112964, "grad_norm": 27.086261749267578, "learning_rate": 4.806146883140805e-05, "loss": 3.6677, "step": 28500 }, { "epoch": 0.41207083355535423, "grad_norm": 27.24114990234375, "learning_rate": 4.793016340815199e-05, "loss": 3.6596, "step": 29000 }, { "epoch": 0.4191755030994121, "grad_norm": 21.950387954711914, "learning_rate": 4.779859484777518e-05, "loss": 3.6288, "step": 29500 }, { "epoch": 0.4262801726434699, "grad_norm": 19.85308074951172, "learning_rate": 4.766702628739836e-05, "loss": 3.6266, "step": 30000 }, { "epoch": 0.4262801726434699, "eval_runtime": 933.7583, "eval_samples_per_second": 150.737, "eval_steps_per_second": 37.684, "step": 30000 }, { "epoch": 0.43338484218752776, "grad_norm": 16.531118392944336, "learning_rate": 4.753545772702155e-05, "loss": 3.65, "step": 30500 }, { "epoch": 0.44048951173158557, "grad_norm": 20.83889389038086, "learning_rate": 4.740388916664474e-05, "loss": 3.6066, "step": 31000 }, { "epoch": 0.44759418127564343, "grad_norm": 17.424823760986328, "learning_rate": 4.727258374338868e-05, "loss": 3.6349, "step": 31500 }, { "epoch": 0.45469885081970124, "grad_norm": 46.09840393066406, "learning_rate": 4.714101518301187e-05, "loss": 3.59, "step": 32000 }, { "epoch": 0.45469885081970124, "eval_runtime": 939.8916, "eval_samples_per_second": 149.753, "eval_steps_per_second": 37.438, "step": 32000 }, { "epoch": 0.4618035203637591, "grad_norm": 18.70945930480957, "learning_rate": 4.700944662263506e-05, "loss": 3.592, "step": 32500 }, { "epoch": 0.4689081899078169, "grad_norm": 20.763996124267578, "learning_rate": 4.687787806225824e-05, "loss": 3.6066, "step": 33000 }, { "epoch": 0.47601285945187477, "grad_norm": 23.204410552978516, "learning_rate": 4.6746572639002185e-05, "loss": 3.6536, "step": 33500 }, { "epoch": 0.4831175289959326, "grad_norm": 18.78313636779785, "learning_rate": 4.661500407862538e-05, "loss": 3.6343, "step": 34000 }, { "epoch": 0.4831175289959326, "eval_runtime": 887.497, "eval_samples_per_second": 158.594, "eval_steps_per_second": 39.649, "step": 34000 }, { "epoch": 0.4902221985399904, "grad_norm": 19.029979705810547, "learning_rate": 4.648343551824856e-05, "loss": 3.6156, "step": 34500 }, { "epoch": 0.49732686808404825, "grad_norm": 22.418338775634766, "learning_rate": 4.635186695787175e-05, "loss": 3.6354, "step": 35000 }, { "epoch": 0.5044315376281061, "grad_norm": 21.143224716186523, "learning_rate": 4.622029839749494e-05, "loss": 3.5547, "step": 35500 }, { "epoch": 0.5115362071721639, "grad_norm": 17.15317153930664, "learning_rate": 4.608925611135963e-05, "loss": 3.5856, "step": 36000 }, { "epoch": 0.5115362071721639, "eval_runtime": 932.1634, "eval_samples_per_second": 150.995, "eval_steps_per_second": 37.749, "step": 36000 }, { "epoch": 0.5186408767162217, "grad_norm": 19.316564559936523, "learning_rate": 4.595768755098282e-05, "loss": 3.5822, "step": 36500 }, { "epoch": 0.5257455462602796, "grad_norm": 44.842105865478516, "learning_rate": 4.582611899060601e-05, "loss": 3.6059, "step": 37000 }, { "epoch": 0.5328502158043374, "grad_norm": 19.391557693481445, "learning_rate": 4.5694550430229194e-05, "loss": 3.6424, "step": 37500 }, { "epoch": 0.5399548853483952, "grad_norm": 17.909786224365234, "learning_rate": 4.5562981869852386e-05, "loss": 3.5643, "step": 38000 }, { "epoch": 0.5399548853483952, "eval_runtime": 906.5953, "eval_samples_per_second": 155.253, "eval_steps_per_second": 38.813, "step": 38000 }, { "epoch": 0.5470595548924531, "grad_norm": 18.2441349029541, "learning_rate": 4.5431413309475565e-05, "loss": 3.5851, "step": 38500 }, { "epoch": 0.5541642244365109, "grad_norm": 28.68638801574707, "learning_rate": 4.529984474909876e-05, "loss": 3.5955, "step": 39000 }, { "epoch": 0.5612688939805688, "grad_norm": 19.42378807067871, "learning_rate": 4.516827618872195e-05, "loss": 3.5832, "step": 39500 }, { "epoch": 0.5683735635246265, "grad_norm": 20.94334602355957, "learning_rate": 4.5036707628345135e-05, "loss": 3.621, "step": 40000 }, { "epoch": 0.5683735635246265, "eval_runtime": 932.967, "eval_samples_per_second": 150.865, "eval_steps_per_second": 37.716, "step": 40000 }, { "epoch": 0.5754782330686844, "grad_norm": 15.059203147888184, "learning_rate": 4.490513906796832e-05, "loss": 3.5537, "step": 40500 }, { "epoch": 0.5825829026127423, "grad_norm": 16.318017959594727, "learning_rate": 4.477383364471226e-05, "loss": 3.5972, "step": 41000 }, { "epoch": 0.5896875721568, "grad_norm": 26.739654541015625, "learning_rate": 4.464226508433545e-05, "loss": 3.5889, "step": 41500 }, { "epoch": 0.5967922417008579, "grad_norm": 17.74175453186035, "learning_rate": 4.451069652395864e-05, "loss": 3.6115, "step": 42000 }, { "epoch": 0.5967922417008579, "eval_runtime": 928.247, "eval_samples_per_second": 151.632, "eval_steps_per_second": 37.908, "step": 42000 }, { "epoch": 0.6038969112449157, "grad_norm": 15.483787536621094, "learning_rate": 4.4379127963581825e-05, "loss": 3.5656, "step": 42500 }, { "epoch": 0.6110015807889736, "grad_norm": 13.45321273803711, "learning_rate": 4.424755940320501e-05, "loss": 3.6075, "step": 43000 }, { "epoch": 0.6181062503330313, "grad_norm": 18.495105743408203, "learning_rate": 4.4115990842828196e-05, "loss": 3.5783, "step": 43500 }, { "epoch": 0.6252109198770892, "grad_norm": 16.545825958251953, "learning_rate": 4.398442228245139e-05, "loss": 3.5468, "step": 44000 }, { "epoch": 0.6252109198770892, "eval_runtime": 924.4432, "eval_samples_per_second": 152.256, "eval_steps_per_second": 38.064, "step": 44000 }, { "epoch": 0.6323155894211471, "grad_norm": 24.224279403686523, "learning_rate": 4.385311685919533e-05, "loss": 3.5977, "step": 44500 }, { "epoch": 0.6394202589652049, "grad_norm": 21.859100341796875, "learning_rate": 4.3721548298818515e-05, "loss": 3.5507, "step": 45000 }, { "epoch": 0.6465249285092627, "grad_norm": 18.18729019165039, "learning_rate": 4.358997973844171e-05, "loss": 3.5176, "step": 45500 }, { "epoch": 0.6536295980533205, "grad_norm": 19.58243179321289, "learning_rate": 4.3458411178064886e-05, "loss": 3.6093, "step": 46000 }, { "epoch": 0.6536295980533205, "eval_runtime": 915.6486, "eval_samples_per_second": 153.718, "eval_steps_per_second": 38.43, "step": 46000 }, { "epoch": 0.6607342675973784, "grad_norm": 18.41427993774414, "learning_rate": 4.332684261768808e-05, "loss": 3.524, "step": 46500 }, { "epoch": 0.6678389371414362, "grad_norm": 17.978282928466797, "learning_rate": 4.319553719443202e-05, "loss": 3.5495, "step": 47000 }, { "epoch": 0.674943606685494, "grad_norm": 20.26556968688965, "learning_rate": 4.3063968634055205e-05, "loss": 3.5603, "step": 47500 }, { "epoch": 0.6820482762295519, "grad_norm": 21.91985511779785, "learning_rate": 4.29324000736784e-05, "loss": 3.5372, "step": 48000 }, { "epoch": 0.6820482762295519, "eval_runtime": 939.9628, "eval_samples_per_second": 149.742, "eval_steps_per_second": 37.436, "step": 48000 }, { "epoch": 0.6891529457736097, "grad_norm": 18.618022918701172, "learning_rate": 4.280083151330158e-05, "loss": 3.5419, "step": 48500 }, { "epoch": 0.6962576153176675, "grad_norm": 20.38291358947754, "learning_rate": 4.266926295292477e-05, "loss": 3.5881, "step": 49000 }, { "epoch": 0.7033622848617254, "grad_norm": 19.36083221435547, "learning_rate": 4.253769439254796e-05, "loss": 3.5256, "step": 49500 }, { "epoch": 0.7104669544057832, "grad_norm": 18.683948516845703, "learning_rate": 4.2406125832171147e-05, "loss": 3.5507, "step": 50000 }, { "epoch": 0.7104669544057832, "eval_runtime": 920.3025, "eval_samples_per_second": 152.941, "eval_steps_per_second": 38.235, "step": 50000 }, { "epoch": 0.7175716239498411, "grad_norm": 15.374938011169434, "learning_rate": 4.227455727179433e-05, "loss": 3.5632, "step": 50500 }, { "epoch": 0.7246762934938988, "grad_norm": 16.976755142211914, "learning_rate": 4.214298871141752e-05, "loss": 3.546, "step": 51000 }, { "epoch": 0.7317809630379567, "grad_norm": 12.849639892578125, "learning_rate": 4.2011683288161466e-05, "loss": 3.5461, "step": 51500 }, { "epoch": 0.7388856325820146, "grad_norm": 19.558488845825195, "learning_rate": 4.188037786490541e-05, "loss": 3.5127, "step": 52000 }, { "epoch": 0.7388856325820146, "eval_runtime": 952.8423, "eval_samples_per_second": 147.718, "eval_steps_per_second": 36.93, "step": 52000 }, { "epoch": 0.7459903021260723, "grad_norm": 19.32866668701172, "learning_rate": 4.174880930452859e-05, "loss": 3.5713, "step": 52500 }, { "epoch": 0.7530949716701302, "grad_norm": 25.51015853881836, "learning_rate": 4.161724074415178e-05, "loss": 3.544, "step": 53000 }, { "epoch": 0.760199641214188, "grad_norm": 23.327672958374023, "learning_rate": 4.1485672183774963e-05, "loss": 3.5789, "step": 53500 }, { "epoch": 0.7673043107582459, "grad_norm": 14.6256685256958, "learning_rate": 4.1354103623398156e-05, "loss": 3.5612, "step": 54000 }, { "epoch": 0.7673043107582459, "eval_runtime": 908.4425, "eval_samples_per_second": 154.938, "eval_steps_per_second": 38.734, "step": 54000 }, { "epoch": 0.7744089803023037, "grad_norm": 13.763489723205566, "learning_rate": 4.12227982001421e-05, "loss": 3.5238, "step": 54500 }, { "epoch": 0.7815136498463615, "grad_norm": 16.628599166870117, "learning_rate": 4.109122963976528e-05, "loss": 3.5555, "step": 55000 }, { "epoch": 0.7886183193904194, "grad_norm": 16.67022132873535, "learning_rate": 4.0959661079388475e-05, "loss": 3.5683, "step": 55500 }, { "epoch": 0.7957229889344772, "grad_norm": 19.672990798950195, "learning_rate": 4.0828092519011654e-05, "loss": 3.4749, "step": 56000 }, { "epoch": 0.7957229889344772, "eval_runtime": 908.7971, "eval_samples_per_second": 154.877, "eval_steps_per_second": 38.719, "step": 56000 }, { "epoch": 0.802827658478535, "grad_norm": 18.64630699157715, "learning_rate": 4.06967870957556e-05, "loss": 3.5202, "step": 56500 }, { "epoch": 0.8099323280225929, "grad_norm": 19.014997482299805, "learning_rate": 4.056521853537879e-05, "loss": 3.514, "step": 57000 }, { "epoch": 0.8170369975666507, "grad_norm": 20.02798843383789, "learning_rate": 4.043364997500197e-05, "loss": 3.5717, "step": 57500 }, { "epoch": 0.8241416671107085, "grad_norm": 13.31798267364502, "learning_rate": 4.0302081414625165e-05, "loss": 3.5485, "step": 58000 }, { "epoch": 0.8241416671107085, "eval_runtime": 920.7476, "eval_samples_per_second": 152.867, "eval_steps_per_second": 38.217, "step": 58000 }, { "epoch": 0.8312463366547663, "grad_norm": 18.563255310058594, "learning_rate": 4.017051285424835e-05, "loss": 3.5037, "step": 58500 }, { "epoch": 0.8383510061988242, "grad_norm": 18.993120193481445, "learning_rate": 4.0038944293871536e-05, "loss": 3.5404, "step": 59000 }, { "epoch": 0.845455675742882, "grad_norm": 23.613004684448242, "learning_rate": 3.990763887061548e-05, "loss": 3.5513, "step": 59500 }, { "epoch": 0.8525603452869398, "grad_norm": 17.62459373474121, "learning_rate": 3.977607031023867e-05, "loss": 3.5172, "step": 60000 }, { "epoch": 0.8525603452869398, "eval_runtime": 947.8413, "eval_samples_per_second": 148.497, "eval_steps_per_second": 37.124, "step": 60000 }, { "epoch": 0.8596650148309977, "grad_norm": 16.220739364624023, "learning_rate": 3.9644501749861855e-05, "loss": 3.5113, "step": 60500 }, { "epoch": 0.8667696843750555, "grad_norm": 21.781057357788086, "learning_rate": 3.951293318948504e-05, "loss": 3.5046, "step": 61000 }, { "epoch": 0.8738743539191134, "grad_norm": 18.97637367248535, "learning_rate": 3.938136462910823e-05, "loss": 3.5144, "step": 61500 }, { "epoch": 0.8809790234631711, "grad_norm": 14.381050109863281, "learning_rate": 3.925005920585217e-05, "loss": 3.4889, "step": 62000 }, { "epoch": 0.8809790234631711, "eval_runtime": 948.6666, "eval_samples_per_second": 148.368, "eval_steps_per_second": 37.092, "step": 62000 }, { "epoch": 0.888083693007229, "grad_norm": 19.86937713623047, "learning_rate": 3.911849064547536e-05, "loss": 3.512, "step": 62500 }, { "epoch": 0.8951883625512869, "grad_norm": 18.57746696472168, "learning_rate": 3.8986922085098545e-05, "loss": 3.5255, "step": 63000 }, { "epoch": 0.9022930320953446, "grad_norm": 23.89400291442871, "learning_rate": 3.885535352472173e-05, "loss": 3.5221, "step": 63500 }, { "epoch": 0.9093977016394025, "grad_norm": 20.12260627746582, "learning_rate": 3.872378496434492e-05, "loss": 3.5573, "step": 64000 }, { "epoch": 0.9093977016394025, "eval_runtime": 957.6922, "eval_samples_per_second": 146.97, "eval_steps_per_second": 36.742, "step": 64000 }, { "epoch": 0.9165023711834603, "grad_norm": 20.893495559692383, "learning_rate": 3.859247954108886e-05, "loss": 3.5487, "step": 64500 }, { "epoch": 0.9236070407275182, "grad_norm": 29.79423713684082, "learning_rate": 3.846091098071205e-05, "loss": 3.5046, "step": 65000 }, { "epoch": 0.930711710271576, "grad_norm": 17.226837158203125, "learning_rate": 3.832934242033524e-05, "loss": 3.5536, "step": 65500 }, { "epoch": 0.9378163798156338, "grad_norm": 33.1728630065918, "learning_rate": 3.819777385995843e-05, "loss": 3.5286, "step": 66000 }, { "epoch": 0.9378163798156338, "eval_runtime": 949.3545, "eval_samples_per_second": 148.261, "eval_steps_per_second": 37.065, "step": 66000 }, { "epoch": 0.9449210493596917, "grad_norm": 19.82408332824707, "learning_rate": 3.806620529958161e-05, "loss": 3.5631, "step": 66500 }, { "epoch": 0.9520257189037495, "grad_norm": 13.408245086669922, "learning_rate": 3.7934899876325554e-05, "loss": 3.4657, "step": 67000 }, { "epoch": 0.9591303884478073, "grad_norm": 16.91946792602539, "learning_rate": 3.780333131594874e-05, "loss": 3.5009, "step": 67500 }, { "epoch": 0.9662350579918652, "grad_norm": 16.867115020751953, "learning_rate": 3.767176275557193e-05, "loss": 3.562, "step": 68000 }, { "epoch": 0.9662350579918652, "eval_runtime": 929.9681, "eval_samples_per_second": 151.351, "eval_steps_per_second": 37.838, "step": 68000 }, { "epoch": 0.973339727535923, "grad_norm": 34.606056213378906, "learning_rate": 3.754019419519512e-05, "loss": 3.5117, "step": 68500 }, { "epoch": 0.9804443970799808, "grad_norm": 16.481327056884766, "learning_rate": 3.74086256348183e-05, "loss": 3.5577, "step": 69000 }, { "epoch": 0.9875490666240386, "grad_norm": 26.235326766967773, "learning_rate": 3.7277057074441496e-05, "loss": 3.5808, "step": 69500 }, { "epoch": 0.9946537361680965, "grad_norm": 25.267833709716797, "learning_rate": 3.714548851406468e-05, "loss": 3.5231, "step": 70000 }, { "epoch": 0.9946537361680965, "eval_runtime": 928.1554, "eval_samples_per_second": 151.647, "eval_steps_per_second": 37.912, "step": 70000 }, { "epoch": 1.0017584057121542, "grad_norm": 17.89148712158203, "learning_rate": 3.7013919953687873e-05, "loss": 3.5025, "step": 70500 }, { "epoch": 1.0088630752562122, "grad_norm": 21.38014793395996, "learning_rate": 3.6882877667552564e-05, "loss": 3.5377, "step": 71000 }, { "epoch": 1.01596774480027, "grad_norm": 17.273181915283203, "learning_rate": 3.675130910717575e-05, "loss": 3.5147, "step": 71500 }, { "epoch": 1.0230724143443277, "grad_norm": 19.168262481689453, "learning_rate": 3.662000368391969e-05, "loss": 3.4739, "step": 72000 }, { "epoch": 1.0230724143443277, "eval_runtime": 921.4794, "eval_samples_per_second": 152.746, "eval_steps_per_second": 38.186, "step": 72000 }, { "epoch": 1.0301770838883857, "grad_norm": 19.674152374267578, "learning_rate": 3.648843512354288e-05, "loss": 3.5197, "step": 72500 }, { "epoch": 1.0372817534324434, "grad_norm": 24.5643367767334, "learning_rate": 3.635686656316607e-05, "loss": 3.4676, "step": 73000 }, { "epoch": 1.0443864229765012, "grad_norm": 16.75780487060547, "learning_rate": 3.6225298002789254e-05, "loss": 3.5216, "step": 73500 }, { "epoch": 1.0514910925205592, "grad_norm": 16.524784088134766, "learning_rate": 3.6093729442412446e-05, "loss": 3.5219, "step": 74000 }, { "epoch": 1.0514910925205592, "eval_runtime": 930.5954, "eval_samples_per_second": 151.249, "eval_steps_per_second": 37.812, "step": 74000 }, { "epoch": 1.058595762064617, "grad_norm": 16.586347579956055, "learning_rate": 3.5962160882035625e-05, "loss": 3.5428, "step": 74500 }, { "epoch": 1.065700431608675, "grad_norm": 17.29988670349121, "learning_rate": 3.583059232165882e-05, "loss": 3.5186, "step": 75000 }, { "epoch": 1.0728051011527326, "grad_norm": 19.46440315246582, "learning_rate": 3.569902376128201e-05, "loss": 3.5, "step": 75500 }, { "epoch": 1.0799097706967904, "grad_norm": 17.439786911010742, "learning_rate": 3.5567455200905195e-05, "loss": 3.5213, "step": 76000 }, { "epoch": 1.0799097706967904, "eval_runtime": 932.7299, "eval_samples_per_second": 150.903, "eval_steps_per_second": 37.726, "step": 76000 }, { "epoch": 1.0870144402408484, "grad_norm": 15.520922660827637, "learning_rate": 3.5436149777649136e-05, "loss": 3.5296, "step": 76500 }, { "epoch": 1.0941191097849061, "grad_norm": 24.197912216186523, "learning_rate": 3.530458121727232e-05, "loss": 3.5079, "step": 77000 }, { "epoch": 1.1012237793289639, "grad_norm": 16.002283096313477, "learning_rate": 3.517301265689551e-05, "loss": 3.5577, "step": 77500 }, { "epoch": 1.1083284488730218, "grad_norm": 23.288475036621094, "learning_rate": 3.50414440965187e-05, "loss": 3.512, "step": 78000 }, { "epoch": 1.1083284488730218, "eval_runtime": 925.8903, "eval_samples_per_second": 152.018, "eval_steps_per_second": 38.005, "step": 78000 }, { "epoch": 1.1154331184170796, "grad_norm": 19.086015701293945, "learning_rate": 3.4909875536141885e-05, "loss": 3.5082, "step": 78500 }, { "epoch": 1.1225377879611376, "grad_norm": 14.41193675994873, "learning_rate": 3.4778570112885826e-05, "loss": 3.479, "step": 79000 }, { "epoch": 1.1296424575051953, "grad_norm": 17.028974533081055, "learning_rate": 3.464700155250901e-05, "loss": 3.4934, "step": 79500 }, { "epoch": 1.136747127049253, "grad_norm": 16.115354537963867, "learning_rate": 3.4515432992132204e-05, "loss": 3.4888, "step": 80000 }, { "epoch": 1.136747127049253, "eval_runtime": 941.5235, "eval_samples_per_second": 149.494, "eval_steps_per_second": 37.373, "step": 80000 }, { "epoch": 1.143851796593311, "grad_norm": 17.744951248168945, "learning_rate": 3.438386443175539e-05, "loss": 3.5219, "step": 80500 }, { "epoch": 1.1509564661373688, "grad_norm": 14.755407333374023, "learning_rate": 3.4252295871378575e-05, "loss": 3.4917, "step": 81000 }, { "epoch": 1.1580611356814265, "grad_norm": 13.909296989440918, "learning_rate": 3.412072731100177e-05, "loss": 3.4889, "step": 81500 }, { "epoch": 1.1651658052254845, "grad_norm": 16.36424446105957, "learning_rate": 3.398915875062495e-05, "loss": 3.5214, "step": 82000 }, { "epoch": 1.1651658052254845, "eval_runtime": 937.9616, "eval_samples_per_second": 150.062, "eval_steps_per_second": 37.515, "step": 82000 }, { "epoch": 1.1722704747695423, "grad_norm": 21.457130432128906, "learning_rate": 3.385759019024814e-05, "loss": 3.499, "step": 82500 }, { "epoch": 1.1793751443136, "grad_norm": 16.613374710083008, "learning_rate": 3.372602162987133e-05, "loss": 3.47, "step": 83000 }, { "epoch": 1.186479813857658, "grad_norm": 20.228662490844727, "learning_rate": 3.3594716206615265e-05, "loss": 3.4882, "step": 83500 }, { "epoch": 1.1935844834017157, "grad_norm": 16.61556625366211, "learning_rate": 3.346314764623846e-05, "loss": 3.4996, "step": 84000 }, { "epoch": 1.1935844834017157, "eval_runtime": 921.749, "eval_samples_per_second": 152.701, "eval_steps_per_second": 38.175, "step": 84000 }, { "epoch": 1.2006891529457735, "grad_norm": 18.255168914794922, "learning_rate": 3.333157908586164e-05, "loss": 3.5136, "step": 84500 }, { "epoch": 1.2077938224898315, "grad_norm": 16.397192001342773, "learning_rate": 3.320001052548483e-05, "loss": 3.4573, "step": 85000 }, { "epoch": 1.2148984920338892, "grad_norm": 23.602087020874023, "learning_rate": 3.306870510222878e-05, "loss": 3.497, "step": 85500 }, { "epoch": 1.2220031615779472, "grad_norm": 23.570209503173828, "learning_rate": 3.293713654185196e-05, "loss": 3.4954, "step": 86000 }, { "epoch": 1.2220031615779472, "eval_runtime": 928.5008, "eval_samples_per_second": 151.591, "eval_steps_per_second": 37.898, "step": 86000 }, { "epoch": 1.229107831122005, "grad_norm": 16.394493103027344, "learning_rate": 3.2805831118595904e-05, "loss": 3.5192, "step": 86500 }, { "epoch": 1.2362125006660627, "grad_norm": 15.098355293273926, "learning_rate": 3.267426255821909e-05, "loss": 3.4441, "step": 87000 }, { "epoch": 1.2433171702101207, "grad_norm": 20.900165557861328, "learning_rate": 3.2542693997842275e-05, "loss": 3.5025, "step": 87500 }, { "epoch": 1.2504218397541784, "grad_norm": 29.24736785888672, "learning_rate": 3.241112543746547e-05, "loss": 3.4749, "step": 88000 }, { "epoch": 1.2504218397541784, "eval_runtime": 925.3923, "eval_samples_per_second": 152.1, "eval_steps_per_second": 38.025, "step": 88000 }, { "epoch": 1.2575265092982364, "grad_norm": 22.28853988647461, "learning_rate": 3.227955687708865e-05, "loss": 3.5224, "step": 88500 }, { "epoch": 1.2646311788422941, "grad_norm": 18.560422897338867, "learning_rate": 3.2148251453832594e-05, "loss": 3.4426, "step": 89000 }, { "epoch": 1.271735848386352, "grad_norm": 22.707122802734375, "learning_rate": 3.201668289345578e-05, "loss": 3.5074, "step": 89500 }, { "epoch": 1.2788405179304099, "grad_norm": 19.690576553344727, "learning_rate": 3.188511433307897e-05, "loss": 3.5421, "step": 90000 }, { "epoch": 1.2788405179304099, "eval_runtime": 923.2428, "eval_samples_per_second": 152.454, "eval_steps_per_second": 38.113, "step": 90000 }, { "epoch": 1.2859451874744676, "grad_norm": 15.056541442871094, "learning_rate": 3.175354577270216e-05, "loss": 3.463, "step": 90500 }, { "epoch": 1.2930498570185254, "grad_norm": 17.073137283325195, "learning_rate": 3.162197721232534e-05, "loss": 3.5104, "step": 91000 }, { "epoch": 1.3001545265625833, "grad_norm": 28.59168243408203, "learning_rate": 3.1490408651948535e-05, "loss": 3.5045, "step": 91500 }, { "epoch": 1.307259196106641, "grad_norm": 20.911029815673828, "learning_rate": 3.135884009157172e-05, "loss": 3.555, "step": 92000 }, { "epoch": 1.307259196106641, "eval_runtime": 929.4581, "eval_samples_per_second": 151.434, "eval_steps_per_second": 37.859, "step": 92000 }, { "epoch": 1.3143638656506988, "grad_norm": 13.504472732543945, "learning_rate": 3.1227271531194906e-05, "loss": 3.4935, "step": 92500 }, { "epoch": 1.3214685351947568, "grad_norm": 16.877857208251953, "learning_rate": 3.1095966107938854e-05, "loss": 3.509, "step": 93000 }, { "epoch": 1.3285732047388146, "grad_norm": 16.726940155029297, "learning_rate": 3.096439754756203e-05, "loss": 3.4867, "step": 93500 }, { "epoch": 1.3356778742828723, "grad_norm": 16.982975006103516, "learning_rate": 3.0832828987185225e-05, "loss": 3.5165, "step": 94000 }, { "epoch": 1.3356778742828723, "eval_runtime": 926.8189, "eval_samples_per_second": 151.866, "eval_steps_per_second": 37.966, "step": 94000 }, { "epoch": 1.3427825438269303, "grad_norm": 23.338428497314453, "learning_rate": 3.070126042680841e-05, "loss": 3.4786, "step": 94500 }, { "epoch": 1.349887213370988, "grad_norm": 17.941577911376953, "learning_rate": 3.056995500355235e-05, "loss": 3.4714, "step": 95000 }, { "epoch": 1.3569918829150458, "grad_norm": 28.672653198242188, "learning_rate": 3.0438386443175544e-05, "loss": 3.5097, "step": 95500 }, { "epoch": 1.3640965524591038, "grad_norm": 18.826953887939453, "learning_rate": 3.0306817882798726e-05, "loss": 3.5081, "step": 96000 }, { "epoch": 1.3640965524591038, "eval_runtime": 938.6101, "eval_samples_per_second": 149.958, "eval_steps_per_second": 37.489, "step": 96000 }, { "epoch": 1.3712012220031615, "grad_norm": 15.427292823791504, "learning_rate": 3.017524932242192e-05, "loss": 3.4842, "step": 96500 }, { "epoch": 1.3783058915472193, "grad_norm": 15.034753799438477, "learning_rate": 3.00436807620451e-05, "loss": 3.5055, "step": 97000 }, { "epoch": 1.3854105610912772, "grad_norm": 16.047800064086914, "learning_rate": 2.9912375338789045e-05, "loss": 3.4855, "step": 97500 }, { "epoch": 1.392515230635335, "grad_norm": 16.472871780395508, "learning_rate": 2.9780806778412234e-05, "loss": 3.4864, "step": 98000 }, { "epoch": 1.392515230635335, "eval_runtime": 923.7428, "eval_samples_per_second": 152.371, "eval_steps_per_second": 38.093, "step": 98000 }, { "epoch": 1.399619900179393, "grad_norm": 15.83158016204834, "learning_rate": 2.964923821803542e-05, "loss": 3.4872, "step": 98500 }, { "epoch": 1.4067245697234507, "grad_norm": 22.84076499938965, "learning_rate": 2.951766965765861e-05, "loss": 3.4432, "step": 99000 }, { "epoch": 1.4138292392675087, "grad_norm": 15.516064643859863, "learning_rate": 2.9386364234402547e-05, "loss": 3.4367, "step": 99500 }, { "epoch": 1.4209339088115664, "grad_norm": 18.551212310791016, "learning_rate": 2.9254795674025735e-05, "loss": 3.4401, "step": 100000 }, { "epoch": 1.4209339088115664, "eval_runtime": 939.3652, "eval_samples_per_second": 149.837, "eval_steps_per_second": 37.459, "step": 100000 }, { "epoch": 1.4280385783556242, "grad_norm": 15.68122673034668, "learning_rate": 2.9123227113648928e-05, "loss": 3.4569, "step": 100500 }, { "epoch": 1.4351432478996822, "grad_norm": 18.92440414428711, "learning_rate": 2.899165855327211e-05, "loss": 3.445, "step": 101000 }, { "epoch": 1.44224791744374, "grad_norm": 21.443042755126953, "learning_rate": 2.8860353130016055e-05, "loss": 3.4473, "step": 101500 }, { "epoch": 1.4493525869877977, "grad_norm": 15.25529670715332, "learning_rate": 2.872878456963924e-05, "loss": 3.4833, "step": 102000 }, { "epoch": 1.4493525869877977, "eval_runtime": 958.8353, "eval_samples_per_second": 146.795, "eval_steps_per_second": 36.699, "step": 102000 }, { "epoch": 1.4564572565318556, "grad_norm": 20.19648551940918, "learning_rate": 2.859721600926243e-05, "loss": 3.4869, "step": 102500 }, { "epoch": 1.4635619260759134, "grad_norm": 20.799802780151367, "learning_rate": 2.8465647448885618e-05, "loss": 3.5262, "step": 103000 }, { "epoch": 1.4706665956199712, "grad_norm": 14.214009284973145, "learning_rate": 2.8334078888508803e-05, "loss": 3.465, "step": 103500 }, { "epoch": 1.4777712651640291, "grad_norm": 14.79015827178955, "learning_rate": 2.8202773465252748e-05, "loss": 3.4532, "step": 104000 }, { "epoch": 1.4777712651640291, "eval_runtime": 938.9838, "eval_samples_per_second": 149.898, "eval_steps_per_second": 37.475, "step": 104000 }, { "epoch": 1.4848759347080869, "grad_norm": 19.22818946838379, "learning_rate": 2.807120490487593e-05, "loss": 3.4953, "step": 104500 }, { "epoch": 1.4919806042521446, "grad_norm": 24.664451599121094, "learning_rate": 2.7939899481619875e-05, "loss": 3.4737, "step": 105000 }, { "epoch": 1.4990852737962026, "grad_norm": 13.840934753417969, "learning_rate": 2.780833092124306e-05, "loss": 3.4858, "step": 105500 }, { "epoch": 1.5061899433402604, "grad_norm": 18.443531036376953, "learning_rate": 2.767676236086625e-05, "loss": 3.4521, "step": 106000 }, { "epoch": 1.5061899433402604, "eval_runtime": 938.3751, "eval_samples_per_second": 149.995, "eval_steps_per_second": 37.499, "step": 106000 }, { "epoch": 1.513294612884318, "grad_norm": 27.919483184814453, "learning_rate": 2.7545193800489438e-05, "loss": 3.4606, "step": 106500 }, { "epoch": 1.520399282428376, "grad_norm": 15.594362258911133, "learning_rate": 2.7413625240112624e-05, "loss": 3.4488, "step": 107000 }, { "epoch": 1.527503951972434, "grad_norm": 15.159219741821289, "learning_rate": 2.7282319816856565e-05, "loss": 3.3977, "step": 107500 }, { "epoch": 1.5346086215164916, "grad_norm": 18.7772159576416, "learning_rate": 2.715075125647975e-05, "loss": 3.4667, "step": 108000 }, { "epoch": 1.5346086215164916, "eval_runtime": 935.6439, "eval_samples_per_second": 150.433, "eval_steps_per_second": 37.608, "step": 108000 }, { "epoch": 1.5417132910605496, "grad_norm": 18.113378524780273, "learning_rate": 2.701918269610294e-05, "loss": 3.4402, "step": 108500 }, { "epoch": 1.5488179606046075, "grad_norm": 23.691232681274414, "learning_rate": 2.6887614135726132e-05, "loss": 3.5135, "step": 109000 }, { "epoch": 1.555922630148665, "grad_norm": 16.19075584411621, "learning_rate": 2.675630871247007e-05, "loss": 3.4753, "step": 109500 }, { "epoch": 1.563027299692723, "grad_norm": 18.575515747070312, "learning_rate": 2.662474015209326e-05, "loss": 3.4945, "step": 110000 }, { "epoch": 1.563027299692723, "eval_runtime": 912.3292, "eval_samples_per_second": 154.278, "eval_steps_per_second": 38.569, "step": 110000 }, { "epoch": 1.570131969236781, "grad_norm": 22.801149368286133, "learning_rate": 2.6493171591716444e-05, "loss": 3.4782, "step": 110500 }, { "epoch": 1.5772366387808388, "grad_norm": 15.964780807495117, "learning_rate": 2.6361603031339633e-05, "loss": 3.4975, "step": 111000 }, { "epoch": 1.5843413083248965, "grad_norm": 15.882137298583984, "learning_rate": 2.6230034470962822e-05, "loss": 3.4472, "step": 111500 }, { "epoch": 1.5914459778689545, "grad_norm": 14.962567329406738, "learning_rate": 2.6098465910586007e-05, "loss": 3.4388, "step": 112000 }, { "epoch": 1.5914459778689545, "eval_runtime": 913.3848, "eval_samples_per_second": 154.099, "eval_steps_per_second": 38.525, "step": 112000 }, { "epoch": 1.5985506474130122, "grad_norm": 21.8176326751709, "learning_rate": 2.596716048732995e-05, "loss": 3.495, "step": 112500 }, { "epoch": 1.60565531695707, "grad_norm": 16.743026733398438, "learning_rate": 2.5835591926953134e-05, "loss": 3.4125, "step": 113000 }, { "epoch": 1.612759986501128, "grad_norm": 18.006900787353516, "learning_rate": 2.5704023366576323e-05, "loss": 3.4638, "step": 113500 }, { "epoch": 1.6198646560451857, "grad_norm": 19.7940731048584, "learning_rate": 2.5572454806199515e-05, "loss": 3.4856, "step": 114000 }, { "epoch": 1.6198646560451857, "eval_runtime": 974.062, "eval_samples_per_second": 144.5, "eval_steps_per_second": 36.125, "step": 114000 }, { "epoch": 1.6269693255892435, "grad_norm": 19.187955856323242, "learning_rate": 2.5440886245822697e-05, "loss": 3.4502, "step": 114500 }, { "epoch": 1.6340739951333014, "grad_norm": 20.042858123779297, "learning_rate": 2.5309317685445886e-05, "loss": 3.517, "step": 115000 }, { "epoch": 1.6411786646773592, "grad_norm": 18.358047485351562, "learning_rate": 2.5178012262189828e-05, "loss": 3.4642, "step": 115500 }, { "epoch": 1.648283334221417, "grad_norm": 15.388800621032715, "learning_rate": 2.5046443701813017e-05, "loss": 3.4721, "step": 116000 }, { "epoch": 1.648283334221417, "eval_runtime": 918.8597, "eval_samples_per_second": 153.181, "eval_steps_per_second": 38.295, "step": 116000 }, { "epoch": 1.655388003765475, "grad_norm": 14.33470630645752, "learning_rate": 2.4914875141436202e-05, "loss": 3.4237, "step": 116500 }, { "epoch": 1.6624926733095327, "grad_norm": 20.042253494262695, "learning_rate": 2.478330658105939e-05, "loss": 3.4401, "step": 117000 }, { "epoch": 1.6695973428535904, "grad_norm": 18.210189819335938, "learning_rate": 2.465173802068258e-05, "loss": 3.459, "step": 117500 }, { "epoch": 1.6767020123976484, "grad_norm": 15.223936080932617, "learning_rate": 2.4520169460305765e-05, "loss": 3.4717, "step": 118000 }, { "epoch": 1.6767020123976484, "eval_runtime": 936.3608, "eval_samples_per_second": 150.318, "eval_steps_per_second": 37.58, "step": 118000 }, { "epoch": 1.6838066819417064, "grad_norm": 17.645105361938477, "learning_rate": 2.4388864037049707e-05, "loss": 3.4599, "step": 118500 }, { "epoch": 1.6909113514857639, "grad_norm": 20.56248664855957, "learning_rate": 2.4257295476672896e-05, "loss": 3.5143, "step": 119000 }, { "epoch": 1.6980160210298219, "grad_norm": 18.7009334564209, "learning_rate": 2.412572691629608e-05, "loss": 3.4491, "step": 119500 }, { "epoch": 1.7051206905738798, "grad_norm": 20.2117977142334, "learning_rate": 2.3994158355919273e-05, "loss": 3.4439, "step": 120000 }, { "epoch": 1.7051206905738798, "eval_runtime": 905.5948, "eval_samples_per_second": 155.425, "eval_steps_per_second": 38.856, "step": 120000 }, { "epoch": 1.7122253601179374, "grad_norm": 16.9014892578125, "learning_rate": 2.386285293266321e-05, "loss": 3.4693, "step": 120500 }, { "epoch": 1.7193300296619953, "grad_norm": 17.105438232421875, "learning_rate": 2.37312843722864e-05, "loss": 3.4436, "step": 121000 }, { "epoch": 1.7264346992060533, "grad_norm": 20.201711654663086, "learning_rate": 2.3599715811909586e-05, "loss": 3.3987, "step": 121500 }, { "epoch": 1.733539368750111, "grad_norm": 16.100046157836914, "learning_rate": 2.3468147251532775e-05, "loss": 3.4672, "step": 122000 }, { "epoch": 1.733539368750111, "eval_runtime": 930.2492, "eval_samples_per_second": 151.306, "eval_steps_per_second": 37.826, "step": 122000 }, { "epoch": 1.7406440382941688, "grad_norm": 21.168506622314453, "learning_rate": 2.3336578691155964e-05, "loss": 3.4675, "step": 122500 }, { "epoch": 1.7477487078382268, "grad_norm": 20.836383819580078, "learning_rate": 2.320501013077915e-05, "loss": 3.4217, "step": 123000 }, { "epoch": 1.7548533773822845, "grad_norm": 18.60804557800293, "learning_rate": 2.3073441570402338e-05, "loss": 3.4592, "step": 123500 }, { "epoch": 1.7619580469263423, "grad_norm": 18.239465713500977, "learning_rate": 2.2941873010025523e-05, "loss": 3.4535, "step": 124000 }, { "epoch": 1.7619580469263423, "eval_runtime": 973.6158, "eval_samples_per_second": 144.566, "eval_steps_per_second": 36.142, "step": 124000 }, { "epoch": 1.7690627164704003, "grad_norm": 14.261148452758789, "learning_rate": 2.2810304449648712e-05, "loss": 3.4712, "step": 124500 }, { "epoch": 1.776167386014458, "grad_norm": 19.631301879882812, "learning_rate": 2.2678999026392657e-05, "loss": 3.4446, "step": 125000 }, { "epoch": 1.7832720555585158, "grad_norm": 16.400991439819336, "learning_rate": 2.2547430466015843e-05, "loss": 3.4876, "step": 125500 }, { "epoch": 1.7903767251025737, "grad_norm": 17.142208099365234, "learning_rate": 2.2415861905639028e-05, "loss": 3.4329, "step": 126000 }, { "epoch": 1.7903767251025737, "eval_runtime": 932.1027, "eval_samples_per_second": 151.005, "eval_steps_per_second": 37.751, "step": 126000 }, { "epoch": 1.7974813946466315, "grad_norm": 24.79787826538086, "learning_rate": 2.228455648238297e-05, "loss": 3.4708, "step": 126500 }, { "epoch": 1.8045860641906892, "grad_norm": 16.074108123779297, "learning_rate": 2.2152987922006158e-05, "loss": 3.4399, "step": 127000 }, { "epoch": 1.8116907337347472, "grad_norm": 13.779826164245605, "learning_rate": 2.2021419361629347e-05, "loss": 3.436, "step": 127500 }, { "epoch": 1.818795403278805, "grad_norm": 23.16010284423828, "learning_rate": 2.1889850801252536e-05, "loss": 3.4112, "step": 128000 }, { "epoch": 1.818795403278805, "eval_runtime": 918.3318, "eval_samples_per_second": 153.269, "eval_steps_per_second": 38.317, "step": 128000 }, { "epoch": 1.8259000728228627, "grad_norm": 14.259251594543457, "learning_rate": 2.175828224087572e-05, "loss": 3.429, "step": 128500 }, { "epoch": 1.8330047423669207, "grad_norm": 13.93868637084961, "learning_rate": 2.1626976817619663e-05, "loss": 3.4684, "step": 129000 }, { "epoch": 1.8401094119109787, "grad_norm": 22.935352325439453, "learning_rate": 2.149540825724285e-05, "loss": 3.4391, "step": 129500 }, { "epoch": 1.8472140814550362, "grad_norm": 19.57620620727539, "learning_rate": 2.136383969686604e-05, "loss": 3.4188, "step": 130000 }, { "epoch": 1.8472140814550362, "eval_runtime": 945.012, "eval_samples_per_second": 148.942, "eval_steps_per_second": 37.236, "step": 130000 }, { "epoch": 1.8543187509990942, "grad_norm": 16.652795791625977, "learning_rate": 2.1232271136489226e-05, "loss": 3.4518, "step": 130500 }, { "epoch": 1.8614234205431521, "grad_norm": 21.736949920654297, "learning_rate": 2.1100702576112412e-05, "loss": 3.4543, "step": 131000 }, { "epoch": 1.8685280900872097, "grad_norm": 17.40629005432129, "learning_rate": 2.09691340157356e-05, "loss": 3.4471, "step": 131500 }, { "epoch": 1.8756327596312676, "grad_norm": 22.435462951660156, "learning_rate": 2.0837565455358786e-05, "loss": 3.4871, "step": 132000 }, { "epoch": 1.8756327596312676, "eval_runtime": 943.7281, "eval_samples_per_second": 149.145, "eval_steps_per_second": 37.286, "step": 132000 }, { "epoch": 1.8827374291753256, "grad_norm": 15.484688758850098, "learning_rate": 2.070599689498198e-05, "loss": 3.4365, "step": 132500 }, { "epoch": 1.8898420987193834, "grad_norm": 20.110355377197266, "learning_rate": 2.057469147172592e-05, "loss": 3.4458, "step": 133000 }, { "epoch": 1.896946768263441, "grad_norm": 27.856857299804688, "learning_rate": 2.0443122911349105e-05, "loss": 3.4394, "step": 133500 }, { "epoch": 1.904051437807499, "grad_norm": 16.6646671295166, "learning_rate": 2.031155435097229e-05, "loss": 3.3948, "step": 134000 }, { "epoch": 1.904051437807499, "eval_runtime": 916.2941, "eval_samples_per_second": 153.61, "eval_steps_per_second": 38.403, "step": 134000 }, { "epoch": 1.9111561073515568, "grad_norm": 17.75814437866211, "learning_rate": 2.017998579059548e-05, "loss": 3.4159, "step": 134500 }, { "epoch": 1.9182607768956146, "grad_norm": 29.82477569580078, "learning_rate": 2.004841723021867e-05, "loss": 3.4183, "step": 135000 }, { "epoch": 1.9253654464396726, "grad_norm": 18.365196228027344, "learning_rate": 1.991711180696261e-05, "loss": 3.4201, "step": 135500 }, { "epoch": 1.9324701159837303, "grad_norm": 19.756254196166992, "learning_rate": 1.9785543246585795e-05, "loss": 3.4536, "step": 136000 }, { "epoch": 1.9324701159837303, "eval_runtime": 927.9317, "eval_samples_per_second": 151.684, "eval_steps_per_second": 37.921, "step": 136000 }, { "epoch": 1.939574785527788, "grad_norm": 24.40082359313965, "learning_rate": 1.9653974686208984e-05, "loss": 3.4828, "step": 136500 }, { "epoch": 1.946679455071846, "grad_norm": 16.37041473388672, "learning_rate": 1.952240612583217e-05, "loss": 3.4541, "step": 137000 }, { "epoch": 1.9537841246159038, "grad_norm": 15.545867919921875, "learning_rate": 1.939110070257611e-05, "loss": 3.4716, "step": 137500 }, { "epoch": 1.9608887941599615, "grad_norm": 27.384632110595703, "learning_rate": 1.9259532142199303e-05, "loss": 3.4382, "step": 138000 }, { "epoch": 1.9608887941599615, "eval_runtime": 910.3209, "eval_samples_per_second": 154.618, "eval_steps_per_second": 38.655, "step": 138000 }, { "epoch": 1.9679934637040195, "grad_norm": 14.11701488494873, "learning_rate": 1.912796358182249e-05, "loss": 3.4248, "step": 138500 }, { "epoch": 1.9750981332480773, "grad_norm": 15.275626182556152, "learning_rate": 1.8996395021445674e-05, "loss": 3.4646, "step": 139000 }, { "epoch": 1.982202802792135, "grad_norm": 15.29659366607666, "learning_rate": 1.8865089598189616e-05, "loss": 3.4473, "step": 139500 }, { "epoch": 1.989307472336193, "grad_norm": 15.644397735595703, "learning_rate": 1.8733521037812805e-05, "loss": 3.4239, "step": 140000 }, { "epoch": 1.989307472336193, "eval_runtime": 938.7255, "eval_samples_per_second": 149.939, "eval_steps_per_second": 37.485, "step": 140000 }, { "epoch": 1.996412141880251, "grad_norm": 19.77324676513672, "learning_rate": 1.8601952477435994e-05, "loss": 3.4572, "step": 140500 }, { "epoch": 2.0035168114243085, "grad_norm": 22.91030502319336, "learning_rate": 1.8470383917059182e-05, "loss": 3.4282, "step": 141000 }, { "epoch": 2.0106214809683665, "grad_norm": 21.323062896728516, "learning_rate": 1.8338815356682368e-05, "loss": 3.4858, "step": 141500 }, { "epoch": 2.0177261505124244, "grad_norm": 15.01187801361084, "learning_rate": 1.8207246796305553e-05, "loss": 3.4014, "step": 142000 }, { "epoch": 2.0177261505124244, "eval_runtime": 887.7808, "eval_samples_per_second": 158.544, "eval_steps_per_second": 39.636, "step": 142000 }, { "epoch": 2.024830820056482, "grad_norm": 17.312917709350586, "learning_rate": 1.8075678235928746e-05, "loss": 3.3528, "step": 142500 }, { "epoch": 2.03193548960054, "grad_norm": 15.480268478393555, "learning_rate": 1.794410967555193e-05, "loss": 3.4759, "step": 143000 }, { "epoch": 2.039040159144598, "grad_norm": 19.289600372314453, "learning_rate": 1.7812804252295873e-05, "loss": 3.4416, "step": 143500 }, { "epoch": 2.0461448286886554, "grad_norm": 13.766185760498047, "learning_rate": 1.7681498829039814e-05, "loss": 3.4511, "step": 144000 }, { "epoch": 2.0461448286886554, "eval_runtime": 942.9255, "eval_samples_per_second": 149.272, "eval_steps_per_second": 37.318, "step": 144000 }, { "epoch": 2.0532494982327134, "grad_norm": 15.767132759094238, "learning_rate": 1.7549930268663e-05, "loss": 3.4972, "step": 144500 }, { "epoch": 2.0603541677767714, "grad_norm": 16.27056121826172, "learning_rate": 1.7418361708286188e-05, "loss": 3.4371, "step": 145000 }, { "epoch": 2.067458837320829, "grad_norm": 19.297935485839844, "learning_rate": 1.7286793147909377e-05, "loss": 3.4419, "step": 145500 }, { "epoch": 2.074563506864887, "grad_norm": 18.82818603515625, "learning_rate": 1.7155224587532566e-05, "loss": 3.4467, "step": 146000 }, { "epoch": 2.074563506864887, "eval_runtime": 921.9813, "eval_samples_per_second": 152.663, "eval_steps_per_second": 38.166, "step": 146000 }, { "epoch": 2.081668176408945, "grad_norm": 16.8089656829834, "learning_rate": 1.702365602715575e-05, "loss": 3.4412, "step": 146500 }, { "epoch": 2.0887728459530024, "grad_norm": 17.762744903564453, "learning_rate": 1.6892087466778937e-05, "loss": 3.4644, "step": 147000 }, { "epoch": 2.0958775154970604, "grad_norm": 17.5285587310791, "learning_rate": 1.6760518906402126e-05, "loss": 3.437, "step": 147500 }, { "epoch": 2.1029821850411183, "grad_norm": 20.605100631713867, "learning_rate": 1.662921348314607e-05, "loss": 3.4551, "step": 148000 }, { "epoch": 2.1029821850411183, "eval_runtime": 902.7522, "eval_samples_per_second": 155.914, "eval_steps_per_second": 38.979, "step": 148000 }, { "epoch": 2.1100868545851763, "grad_norm": 17.58387565612793, "learning_rate": 1.6497644922769256e-05, "loss": 3.4763, "step": 148500 }, { "epoch": 2.117191524129234, "grad_norm": 20.26319122314453, "learning_rate": 1.6366076362392445e-05, "loss": 3.4081, "step": 149000 }, { "epoch": 2.124296193673292, "grad_norm": 12.938958168029785, "learning_rate": 1.623450780201563e-05, "loss": 3.4738, "step": 149500 }, { "epoch": 2.13140086321735, "grad_norm": 18.24934196472168, "learning_rate": 1.6102939241638816e-05, "loss": 3.4015, "step": 150000 }, { "epoch": 2.13140086321735, "eval_runtime": 952.2281, "eval_samples_per_second": 147.813, "eval_steps_per_second": 36.953, "step": 150000 }, { "epoch": 2.1385055327614073, "grad_norm": 17.659948348999023, "learning_rate": 1.597137068126201e-05, "loss": 3.4798, "step": 150500 }, { "epoch": 2.1456102023054653, "grad_norm": 16.74921226501465, "learning_rate": 1.5839802120885194e-05, "loss": 3.4662, "step": 151000 }, { "epoch": 2.1527148718495233, "grad_norm": 18.972862243652344, "learning_rate": 1.5708496697629135e-05, "loss": 3.4461, "step": 151500 }, { "epoch": 2.159819541393581, "grad_norm": 17.463333129882812, "learning_rate": 1.557692813725232e-05, "loss": 3.4509, "step": 152000 }, { "epoch": 2.159819541393581, "eval_runtime": 952.341, "eval_samples_per_second": 147.796, "eval_steps_per_second": 36.949, "step": 152000 }, { "epoch": 2.1669242109376388, "grad_norm": 17.932537078857422, "learning_rate": 1.544535957687551e-05, "loss": 3.4188, "step": 152500 }, { "epoch": 2.1740288804816967, "grad_norm": 18.25333023071289, "learning_rate": 1.53137910164987e-05, "loss": 3.4275, "step": 153000 }, { "epoch": 2.1811335500257543, "grad_norm": 15.192218780517578, "learning_rate": 1.5182222456121886e-05, "loss": 3.4497, "step": 153500 }, { "epoch": 2.1882382195698122, "grad_norm": 59.057674407958984, "learning_rate": 1.5050653895745073e-05, "loss": 3.4341, "step": 154000 }, { "epoch": 2.1882382195698122, "eval_runtime": 907.5029, "eval_samples_per_second": 155.098, "eval_steps_per_second": 38.775, "step": 154000 }, { "epoch": 2.19534288911387, "grad_norm": 17.72281265258789, "learning_rate": 1.491908533536826e-05, "loss": 3.4777, "step": 154500 }, { "epoch": 2.2024475586579277, "grad_norm": 17.717147827148438, "learning_rate": 1.478751677499145e-05, "loss": 3.4275, "step": 155000 }, { "epoch": 2.2095522282019857, "grad_norm": 20.065603256225586, "learning_rate": 1.4655948214614636e-05, "loss": 3.3751, "step": 155500 }, { "epoch": 2.2166568977460437, "grad_norm": 19.137521743774414, "learning_rate": 1.4524642791358578e-05, "loss": 3.4704, "step": 156000 }, { "epoch": 2.2166568977460437, "eval_runtime": 923.9445, "eval_samples_per_second": 152.338, "eval_steps_per_second": 38.085, "step": 156000 }, { "epoch": 2.223761567290101, "grad_norm": 28.932388305664062, "learning_rate": 1.4393074230981765e-05, "loss": 3.4592, "step": 156500 }, { "epoch": 2.230866236834159, "grad_norm": 18.90865135192871, "learning_rate": 1.4261505670604952e-05, "loss": 3.4281, "step": 157000 }, { "epoch": 2.237970906378217, "grad_norm": 14.286882400512695, "learning_rate": 1.4129937110228143e-05, "loss": 3.4169, "step": 157500 }, { "epoch": 2.245075575922275, "grad_norm": 25.19231605529785, "learning_rate": 1.3998631686972082e-05, "loss": 3.4514, "step": 158000 }, { "epoch": 2.245075575922275, "eval_runtime": 955.1982, "eval_samples_per_second": 147.354, "eval_steps_per_second": 36.838, "step": 158000 }, { "epoch": 2.2521802454663327, "grad_norm": 15.808695793151855, "learning_rate": 1.386706312659527e-05, "loss": 3.4476, "step": 158500 }, { "epoch": 2.2592849150103906, "grad_norm": 20.826406478881836, "learning_rate": 1.3735494566218457e-05, "loss": 3.4425, "step": 159000 }, { "epoch": 2.2663895845544486, "grad_norm": 16.025421142578125, "learning_rate": 1.3603926005841644e-05, "loss": 3.4537, "step": 159500 }, { "epoch": 2.273494254098506, "grad_norm": 23.02109146118164, "learning_rate": 1.3472620582585585e-05, "loss": 3.4596, "step": 160000 }, { "epoch": 2.273494254098506, "eval_runtime": 994.6152, "eval_samples_per_second": 141.514, "eval_steps_per_second": 35.379, "step": 160000 }, { "epoch": 2.280598923642564, "grad_norm": 24.307044982910156, "learning_rate": 1.3341052022208774e-05, "loss": 3.4252, "step": 160500 }, { "epoch": 2.287703593186622, "grad_norm": 18.153467178344727, "learning_rate": 1.3209483461831961e-05, "loss": 3.412, "step": 161000 }, { "epoch": 2.2948082627306796, "grad_norm": 22.00274658203125, "learning_rate": 1.3077914901455148e-05, "loss": 3.4418, "step": 161500 }, { "epoch": 2.3019129322747376, "grad_norm": 14.408377647399902, "learning_rate": 1.294660947819909e-05, "loss": 3.4745, "step": 162000 }, { "epoch": 2.3019129322747376, "eval_runtime": 957.1425, "eval_samples_per_second": 147.054, "eval_steps_per_second": 36.764, "step": 162000 }, { "epoch": 2.3090176018187956, "grad_norm": 17.56698989868164, "learning_rate": 1.2815040917822277e-05, "loss": 3.4048, "step": 162500 }, { "epoch": 2.316122271362853, "grad_norm": 14.062944412231445, "learning_rate": 1.2683472357445468e-05, "loss": 3.4323, "step": 163000 }, { "epoch": 2.323226940906911, "grad_norm": 16.818817138671875, "learning_rate": 1.2551903797068653e-05, "loss": 3.4252, "step": 163500 }, { "epoch": 2.330331610450969, "grad_norm": 29.905399322509766, "learning_rate": 1.242033523669184e-05, "loss": 3.4526, "step": 164000 }, { "epoch": 2.330331610450969, "eval_runtime": 942.1268, "eval_samples_per_second": 149.398, "eval_steps_per_second": 37.35, "step": 164000 }, { "epoch": 2.3374362799950266, "grad_norm": 15.426836967468262, "learning_rate": 1.2288766676315027e-05, "loss": 3.3833, "step": 164500 }, { "epoch": 2.3445409495390845, "grad_norm": 17.668203353881836, "learning_rate": 1.2157198115938216e-05, "loss": 3.4454, "step": 165000 }, { "epoch": 2.3516456190831425, "grad_norm": 19.938480377197266, "learning_rate": 1.2025629555561404e-05, "loss": 3.4777, "step": 165500 }, { "epoch": 2.3587502886272, "grad_norm": 25.68340492248535, "learning_rate": 1.1894324132305345e-05, "loss": 3.4294, "step": 166000 }, { "epoch": 2.3587502886272, "eval_runtime": 922.2117, "eval_samples_per_second": 152.624, "eval_steps_per_second": 38.156, "step": 166000 }, { "epoch": 2.365854958171258, "grad_norm": 27.432424545288086, "learning_rate": 1.1762755571928532e-05, "loss": 3.4314, "step": 166500 }, { "epoch": 2.372959627715316, "grad_norm": 16.082857131958008, "learning_rate": 1.163118701155172e-05, "loss": 3.4351, "step": 167000 }, { "epoch": 2.3800642972593735, "grad_norm": 22.185914993286133, "learning_rate": 1.1499618451174908e-05, "loss": 3.461, "step": 167500 }, { "epoch": 2.3871689668034315, "grad_norm": 24.197967529296875, "learning_rate": 1.1368049890798095e-05, "loss": 3.4533, "step": 168000 }, { "epoch": 2.3871689668034315, "eval_runtime": 895.4635, "eval_samples_per_second": 157.183, "eval_steps_per_second": 39.296, "step": 168000 }, { "epoch": 2.3942736363474895, "grad_norm": 17.674205780029297, "learning_rate": 1.1236744467542037e-05, "loss": 3.4427, "step": 168500 }, { "epoch": 2.401378305891547, "grad_norm": 18.028247833251953, "learning_rate": 1.1105175907165224e-05, "loss": 3.4074, "step": 169000 }, { "epoch": 2.408482975435605, "grad_norm": 22.633684158325195, "learning_rate": 1.0973607346788411e-05, "loss": 3.4411, "step": 169500 }, { "epoch": 2.415587644979663, "grad_norm": 20.172569274902344, "learning_rate": 1.08420387864116e-05, "loss": 3.4517, "step": 170000 }, { "epoch": 2.415587644979663, "eval_runtime": 934.155, "eval_samples_per_second": 150.673, "eval_steps_per_second": 37.668, "step": 170000 }, { "epoch": 2.4226923145237205, "grad_norm": 20.472396850585938, "learning_rate": 1.0710733363155541e-05, "loss": 3.4382, "step": 170500 }, { "epoch": 2.4297969840677784, "grad_norm": 18.737754821777344, "learning_rate": 1.0579164802778729e-05, "loss": 3.3906, "step": 171000 }, { "epoch": 2.4369016536118364, "grad_norm": 25.360733032226562, "learning_rate": 1.0447596242401916e-05, "loss": 3.4101, "step": 171500 }, { "epoch": 2.4440063231558944, "grad_norm": 17.960433959960938, "learning_rate": 1.0316027682025103e-05, "loss": 3.4239, "step": 172000 }, { "epoch": 2.4440063231558944, "eval_runtime": 902.2, "eval_samples_per_second": 156.01, "eval_steps_per_second": 39.002, "step": 172000 }, { "epoch": 2.451110992699952, "grad_norm": 14.151226997375488, "learning_rate": 1.0184722258769044e-05, "loss": 3.4173, "step": 172500 }, { "epoch": 2.45821566224401, "grad_norm": 14.142123222351074, "learning_rate": 1.0053153698392233e-05, "loss": 3.4156, "step": 173000 }, { "epoch": 2.465320331788068, "grad_norm": 21.70526695251465, "learning_rate": 9.921848275136174e-06, "loss": 3.458, "step": 173500 }, { "epoch": 2.4724250013321254, "grad_norm": 18.53485870361328, "learning_rate": 9.790279714759362e-06, "loss": 3.447, "step": 174000 }, { "epoch": 2.4724250013321254, "eval_runtime": 938.1633, "eval_samples_per_second": 150.029, "eval_steps_per_second": 37.507, "step": 174000 }, { "epoch": 2.4795296708761834, "grad_norm": 18.21489143371582, "learning_rate": 9.658711154382549e-06, "loss": 3.3984, "step": 174500 }, { "epoch": 2.4866343404202413, "grad_norm": 16.120616912841797, "learning_rate": 9.527142594005736e-06, "loss": 3.4071, "step": 175000 }, { "epoch": 2.493739009964299, "grad_norm": 15.231195449829102, "learning_rate": 9.395574033628925e-06, "loss": 3.3855, "step": 175500 }, { "epoch": 2.500843679508357, "grad_norm": 26.148868560791016, "learning_rate": 9.264005473252112e-06, "loss": 3.3718, "step": 176000 }, { "epoch": 2.500843679508357, "eval_runtime": 931.3581, "eval_samples_per_second": 151.126, "eval_steps_per_second": 37.781, "step": 176000 }, { "epoch": 2.507948349052415, "grad_norm": 22.703458786010742, "learning_rate": 9.1324369128753e-06, "loss": 3.4483, "step": 176500 }, { "epoch": 2.515053018596473, "grad_norm": 24.028915405273438, "learning_rate": 9.000868352498487e-06, "loss": 3.4646, "step": 177000 }, { "epoch": 2.5221576881405303, "grad_norm": 15.914198875427246, "learning_rate": 8.869562929242428e-06, "loss": 3.3819, "step": 177500 }, { "epoch": 2.5292623576845883, "grad_norm": 20.722768783569336, "learning_rate": 8.737994368865617e-06, "loss": 3.4321, "step": 178000 }, { "epoch": 2.5292623576845883, "eval_runtime": 931.1256, "eval_samples_per_second": 151.163, "eval_steps_per_second": 37.791, "step": 178000 }, { "epoch": 2.5363670272286463, "grad_norm": 16.944055557250977, "learning_rate": 8.606425808488804e-06, "loss": 3.4277, "step": 178500 }, { "epoch": 2.543471696772704, "grad_norm": 19.507802963256836, "learning_rate": 8.474857248111991e-06, "loss": 3.4652, "step": 179000 }, { "epoch": 2.5505763663167618, "grad_norm": 17.919048309326172, "learning_rate": 8.343288687735178e-06, "loss": 3.4398, "step": 179500 }, { "epoch": 2.5576810358608197, "grad_norm": 14.281726837158203, "learning_rate": 8.211720127358367e-06, "loss": 3.3992, "step": 180000 }, { "epoch": 2.5576810358608197, "eval_runtime": 944.9306, "eval_samples_per_second": 148.955, "eval_steps_per_second": 37.239, "step": 180000 }, { "epoch": 2.5647857054048773, "grad_norm": 14.9695463180542, "learning_rate": 8.080151566981554e-06, "loss": 3.4413, "step": 180500 }, { "epoch": 2.5718903749489352, "grad_norm": 16.833791732788086, "learning_rate": 7.948583006604742e-06, "loss": 3.4267, "step": 181000 }, { "epoch": 2.578995044492993, "grad_norm": 15.470376014709473, "learning_rate": 7.81701444622793e-06, "loss": 3.4481, "step": 181500 }, { "epoch": 2.5860997140370507, "grad_norm": 21.1016845703125, "learning_rate": 7.68570902297187e-06, "loss": 3.4727, "step": 182000 }, { "epoch": 2.5860997140370507, "eval_runtime": 954.808, "eval_samples_per_second": 147.414, "eval_steps_per_second": 36.853, "step": 182000 }, { "epoch": 2.5932043835811087, "grad_norm": 21.787841796875, "learning_rate": 7.554140462595059e-06, "loss": 3.4596, "step": 182500 }, { "epoch": 2.6003090531251667, "grad_norm": 17.705780029296875, "learning_rate": 7.422571902218246e-06, "loss": 3.4334, "step": 183000 }, { "epoch": 2.607413722669224, "grad_norm": 24.78368377685547, "learning_rate": 7.2910033418414335e-06, "loss": 3.4673, "step": 183500 }, { "epoch": 2.614518392213282, "grad_norm": 17.67124366760254, "learning_rate": 7.159697918585376e-06, "loss": 3.398, "step": 184000 }, { "epoch": 2.614518392213282, "eval_runtime": 950.0509, "eval_samples_per_second": 148.152, "eval_steps_per_second": 37.038, "step": 184000 }, { "epoch": 2.62162306175734, "grad_norm": 15.1649808883667, "learning_rate": 7.028129358208563e-06, "loss": 3.4236, "step": 184500 }, { "epoch": 2.6287277313013977, "grad_norm": 18.517230987548828, "learning_rate": 6.89656079783175e-06, "loss": 3.3925, "step": 185000 }, { "epoch": 2.6358324008454557, "grad_norm": 15.220212936401367, "learning_rate": 6.764992237454938e-06, "loss": 3.3708, "step": 185500 }, { "epoch": 2.6429370703895136, "grad_norm": 16.052600860595703, "learning_rate": 6.6336868141988795e-06, "loss": 3.4095, "step": 186000 }, { "epoch": 2.6429370703895136, "eval_runtime": 902.4427, "eval_samples_per_second": 155.968, "eval_steps_per_second": 38.992, "step": 186000 }, { "epoch": 2.650041739933571, "grad_norm": 17.24014663696289, "learning_rate": 6.50238139094282e-06, "loss": 3.3694, "step": 186500 }, { "epoch": 2.657146409477629, "grad_norm": 15.43420696258545, "learning_rate": 6.370812830566009e-06, "loss": 3.4659, "step": 187000 }, { "epoch": 2.664251079021687, "grad_norm": 19.86100959777832, "learning_rate": 6.239507407309949e-06, "loss": 3.4001, "step": 187500 }, { "epoch": 2.6713557485657446, "grad_norm": 15.499372482299805, "learning_rate": 6.107938846933137e-06, "loss": 3.4005, "step": 188000 }, { "epoch": 2.6713557485657446, "eval_runtime": 935.9412, "eval_samples_per_second": 150.386, "eval_steps_per_second": 37.596, "step": 188000 }, { "epoch": 2.6784604181098026, "grad_norm": 18.945688247680664, "learning_rate": 5.976370286556325e-06, "loss": 3.4568, "step": 188500 }, { "epoch": 2.6855650876538606, "grad_norm": 23.000934600830078, "learning_rate": 5.844801726179512e-06, "loss": 3.4078, "step": 189000 }, { "epoch": 2.692669757197918, "grad_norm": 16.832963943481445, "learning_rate": 5.7132331658027e-06, "loss": 3.4144, "step": 189500 }, { "epoch": 2.699774426741976, "grad_norm": 21.619384765625, "learning_rate": 5.581664605425888e-06, "loss": 3.4234, "step": 190000 }, { "epoch": 2.699774426741976, "eval_runtime": 977.3575, "eval_samples_per_second": 144.013, "eval_steps_per_second": 36.003, "step": 190000 }, { "epoch": 2.706879096286034, "grad_norm": 34.39583206176758, "learning_rate": 5.450096045049075e-06, "loss": 3.4315, "step": 190500 }, { "epoch": 2.7139837658300916, "grad_norm": 18.354318618774414, "learning_rate": 5.318527484672263e-06, "loss": 3.4279, "step": 191000 }, { "epoch": 2.7210884353741496, "grad_norm": 24.28253173828125, "learning_rate": 5.18695892429545e-06, "loss": 3.4595, "step": 191500 }, { "epoch": 2.7281931049182075, "grad_norm": 23.663331985473633, "learning_rate": 5.055390363918638e-06, "loss": 3.3998, "step": 192000 }, { "epoch": 2.7281931049182075, "eval_runtime": 916.9506, "eval_samples_per_second": 153.5, "eval_steps_per_second": 38.375, "step": 192000 }, { "epoch": 2.735297774462265, "grad_norm": 22.13831329345703, "learning_rate": 4.923821803541826e-06, "loss": 3.4417, "step": 192500 }, { "epoch": 2.742402444006323, "grad_norm": 19.08433723449707, "learning_rate": 4.792253243165014e-06, "loss": 3.4359, "step": 193000 }, { "epoch": 2.749507113550381, "grad_norm": 14.219480514526367, "learning_rate": 4.660947819908955e-06, "loss": 3.3982, "step": 193500 }, { "epoch": 2.7566117830944386, "grad_norm": 16.88738250732422, "learning_rate": 4.529379259532143e-06, "loss": 3.3999, "step": 194000 }, { "epoch": 2.7566117830944386, "eval_runtime": 929.4031, "eval_samples_per_second": 151.443, "eval_steps_per_second": 37.861, "step": 194000 }, { "epoch": 2.7637164526384965, "grad_norm": 24.45867347717285, "learning_rate": 4.39781069915533e-06, "loss": 3.4601, "step": 194500 }, { "epoch": 2.7708211221825545, "grad_norm": 18.18239402770996, "learning_rate": 4.266242138778518e-06, "loss": 3.4752, "step": 195000 }, { "epoch": 2.7779257917266125, "grad_norm": 25.724437713623047, "learning_rate": 4.134936715522459e-06, "loss": 3.4426, "step": 195500 }, { "epoch": 2.78503046127067, "grad_norm": 22.24292755126953, "learning_rate": 4.003368155145647e-06, "loss": 3.422, "step": 196000 }, { "epoch": 2.78503046127067, "eval_runtime": 953.0692, "eval_samples_per_second": 147.683, "eval_steps_per_second": 36.921, "step": 196000 }, { "epoch": 2.792135130814728, "grad_norm": 14.148300170898438, "learning_rate": 3.871799594768835e-06, "loss": 3.4318, "step": 196500 }, { "epoch": 2.799239800358786, "grad_norm": 20.823726654052734, "learning_rate": 3.740231034392022e-06, "loss": 3.4268, "step": 197000 }, { "epoch": 2.806344469902844, "grad_norm": 17.67780876159668, "learning_rate": 3.6089256111359633e-06, "loss": 3.4177, "step": 197500 }, { "epoch": 2.8134491394469014, "grad_norm": 15.137374877929688, "learning_rate": 3.477357050759151e-06, "loss": 3.4456, "step": 198000 }, { "epoch": 2.8134491394469014, "eval_runtime": 937.5641, "eval_samples_per_second": 150.125, "eval_steps_per_second": 37.531, "step": 198000 }, { "epoch": 2.8205538089909594, "grad_norm": 19.98059844970703, "learning_rate": 3.3457884903823386e-06, "loss": 3.4413, "step": 198500 }, { "epoch": 2.8276584785350174, "grad_norm": 21.89630126953125, "learning_rate": 3.2144830671262795e-06, "loss": 3.4022, "step": 199000 }, { "epoch": 2.834763148079075, "grad_norm": 14.519454956054688, "learning_rate": 3.0829145067494675e-06, "loss": 3.4555, "step": 199500 }, { "epoch": 2.841867817623133, "grad_norm": 16.703575134277344, "learning_rate": 2.951345946372655e-06, "loss": 3.4092, "step": 200000 }, { "epoch": 2.841867817623133, "eval_runtime": 953.5002, "eval_samples_per_second": 147.616, "eval_steps_per_second": 36.904, "step": 200000 }, { "epoch": 2.848972487167191, "grad_norm": 27.454334259033203, "learning_rate": 2.8197773859958423e-06, "loss": 3.418, "step": 200500 }, { "epoch": 2.8560771567112484, "grad_norm": 23.255996704101562, "learning_rate": 2.68820882561903e-06, "loss": 3.3987, "step": 201000 }, { "epoch": 2.8631818262553064, "grad_norm": 15.948774337768555, "learning_rate": 2.556640265242218e-06, "loss": 3.4912, "step": 201500 }, { "epoch": 2.8702864957993643, "grad_norm": 20.497947692871094, "learning_rate": 2.4250717048654057e-06, "loss": 3.448, "step": 202000 }, { "epoch": 2.8702864957993643, "eval_runtime": 911.0819, "eval_samples_per_second": 154.489, "eval_steps_per_second": 38.622, "step": 202000 }, { "epoch": 2.877391165343422, "grad_norm": 15.809967994689941, "learning_rate": 2.2937662816093465e-06, "loss": 3.4309, "step": 202500 }, { "epoch": 2.88449583488748, "grad_norm": 18.75983238220215, "learning_rate": 2.162197721232534e-06, "loss": 3.4832, "step": 203000 }, { "epoch": 2.891600504431538, "grad_norm": 18.350236892700195, "learning_rate": 2.030629160855722e-06, "loss": 3.4685, "step": 203500 }, { "epoch": 2.8987051739755954, "grad_norm": 22.548627853393555, "learning_rate": 1.8990606004789096e-06, "loss": 3.4514, "step": 204000 }, { "epoch": 2.8987051739755954, "eval_runtime": 907.2069, "eval_samples_per_second": 155.149, "eval_steps_per_second": 38.787, "step": 204000 }, { "epoch": 2.9058098435196533, "grad_norm": 22.571067810058594, "learning_rate": 1.7674920401020975e-06, "loss": 3.383, "step": 204500 }, { "epoch": 2.9129145130637113, "grad_norm": 20.21802520751953, "learning_rate": 1.635923479725285e-06, "loss": 3.4761, "step": 205000 }, { "epoch": 2.920019182607769, "grad_norm": 22.196605682373047, "learning_rate": 1.5043549193484725e-06, "loss": 3.4293, "step": 205500 }, { "epoch": 2.927123852151827, "grad_norm": 20.584014892578125, "learning_rate": 1.3727863589716602e-06, "loss": 3.4388, "step": 206000 }, { "epoch": 2.927123852151827, "eval_runtime": 935.9699, "eval_samples_per_second": 150.381, "eval_steps_per_second": 37.595, "step": 206000 }, { "epoch": 2.9342285216958848, "grad_norm": 21.364656448364258, "learning_rate": 1.2414809357156015e-06, "loss": 3.3946, "step": 206500 }, { "epoch": 2.9413331912399423, "grad_norm": 19.831722259521484, "learning_rate": 1.109912375338789e-06, "loss": 3.389, "step": 207000 }, { "epoch": 2.9484378607840003, "grad_norm": 61.1855583190918, "learning_rate": 9.783438149619767e-07, "loss": 3.4133, "step": 207500 }, { "epoch": 2.9555425303280582, "grad_norm": 22.293066024780273, "learning_rate": 8.467752545851645e-07, "loss": 3.4065, "step": 208000 }, { "epoch": 2.9555425303280582, "eval_runtime": 952.4375, "eval_samples_per_second": 147.781, "eval_steps_per_second": 36.945, "step": 208000 }, { "epoch": 2.9626471998721158, "grad_norm": 22.37586212158203, "learning_rate": 7.152066942083521e-07, "loss": 3.4225, "step": 208500 }, { "epoch": 2.9697518694161738, "grad_norm": 22.698583602905273, "learning_rate": 5.839012709522933e-07, "loss": 3.4399, "step": 209000 }, { "epoch": 2.9768565389602317, "grad_norm": 19.18611717224121, "learning_rate": 4.523327105754809e-07, "loss": 3.3873, "step": 209500 }, { "epoch": 2.9839612085042893, "grad_norm": 20.257211685180664, "learning_rate": 3.2076415019866855e-07, "loss": 3.4055, "step": 210000 }, { "epoch": 2.9839612085042893, "eval_runtime": 935.2702, "eval_samples_per_second": 150.493, "eval_steps_per_second": 37.623, "step": 210000 } ], "logging_steps": 500, "max_steps": 211128, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 2000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 4.5835319333291544e+16, "train_batch_size": 4, "trial_name": null, "trial_params": null }