Files
math-1b-sft-numinamath-bs51…/trainer_state.json
ModelHub XC 6f0dcbefc2 初始化项目,由ModelHub XC社区提供模型
Model: pre-to-post-olmo/math-1b-sft-numinamath-bs512-from-step45000
Source: Original Platform
2026-07-19 16:20:44 +08:00

1213 lines
30 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 3358,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.005956813104988831,
"grad_norm": 2.680657148361206,
"learning_rate": 1.8811881188118813e-06,
"loss": 1.8061,
"step": 20
},
{
"epoch": 0.011913626209977662,
"grad_norm": 1.1645660400390625,
"learning_rate": 3.861386138613862e-06,
"loss": 1.5899,
"step": 40
},
{
"epoch": 0.017870439314966492,
"grad_norm": 0.4798089861869812,
"learning_rate": 5.841584158415842e-06,
"loss": 1.1208,
"step": 60
},
{
"epoch": 0.023827252419955324,
"grad_norm": 0.2725222706794739,
"learning_rate": 7.821782178217822e-06,
"loss": 0.8586,
"step": 80
},
{
"epoch": 0.029784065524944156,
"grad_norm": 0.18464167416095734,
"learning_rate": 9.801980198019802e-06,
"loss": 0.7286,
"step": 100
},
{
"epoch": 0.035740878629932984,
"grad_norm": 0.1603326052427292,
"learning_rate": 9.99924640442892e-06,
"loss": 0.6877,
"step": 120
},
{
"epoch": 0.04169769173492182,
"grad_norm": 0.1472804993391037,
"learning_rate": 9.99664167434275e-06,
"loss": 0.6602,
"step": 140
},
{
"epoch": 0.04765450483991065,
"grad_norm": 0.13646265864372253,
"learning_rate": 9.99217747519571e-06,
"loss": 0.6395,
"step": 160
},
{
"epoch": 0.05361131794489948,
"grad_norm": 0.140568345785141,
"learning_rate": 9.98585546831169e-06,
"loss": 0.6358,
"step": 180
},
{
"epoch": 0.05956813104988831,
"grad_norm": 0.12733544409275055,
"learning_rate": 9.977678006386153e-06,
"loss": 0.6139,
"step": 200
},
{
"epoch": 0.06552494415487714,
"grad_norm": 0.1283230483531952,
"learning_rate": 9.967648132610596e-06,
"loss": 0.6122,
"step": 220
},
{
"epoch": 0.07148175725986597,
"grad_norm": 0.1346861720085144,
"learning_rate": 9.95576957954004e-06,
"loss": 0.6081,
"step": 240
},
{
"epoch": 0.0774385703648548,
"grad_norm": 0.12492821365594864,
"learning_rate": 9.942046767703982e-06,
"loss": 0.605,
"step": 260
},
{
"epoch": 0.08339538346984364,
"grad_norm": 0.1220928356051445,
"learning_rate": 9.926484803961332e-06,
"loss": 0.5975,
"step": 280
},
{
"epoch": 0.08935219657483247,
"grad_norm": 0.1256948560476303,
"learning_rate": 9.909089479599916e-06,
"loss": 0.5973,
"step": 300
},
{
"epoch": 0.0953090096798213,
"grad_norm": 0.12339841574430466,
"learning_rate": 9.88986726818129e-06,
"loss": 0.5983,
"step": 320
},
{
"epoch": 0.10126582278481013,
"grad_norm": 0.1248016506433487,
"learning_rate": 9.868825323131646e-06,
"loss": 0.5841,
"step": 340
},
{
"epoch": 0.10722263588979895,
"grad_norm": 0.12367212027311325,
"learning_rate": 9.8459714750797e-06,
"loss": 0.584,
"step": 360
},
{
"epoch": 0.11317944899478778,
"grad_norm": 0.12310680001974106,
"learning_rate": 9.821314228942587e-06,
"loss": 0.5916,
"step": 380
},
{
"epoch": 0.11913626209977662,
"grad_norm": 0.12211985141038895,
"learning_rate": 9.794862760760798e-06,
"loss": 0.5825,
"step": 400
},
{
"epoch": 0.12509307520476545,
"grad_norm": 0.11687470227479935,
"learning_rate": 9.76662691428337e-06,
"loss": 0.5859,
"step": 420
},
{
"epoch": 0.13104988830975428,
"grad_norm": 0.11877286434173584,
"learning_rate": 9.736617197304615e-06,
"loss": 0.5757,
"step": 440
},
{
"epoch": 0.1370067014147431,
"grad_norm": 0.11676610261201859,
"learning_rate": 9.704844777753686e-06,
"loss": 0.5706,
"step": 460
},
{
"epoch": 0.14296351451973194,
"grad_norm": 0.12015936523675919,
"learning_rate": 9.671321479538522e-06,
"loss": 0.5713,
"step": 480
},
{
"epoch": 0.14892032762472077,
"grad_norm": 0.12606967985630035,
"learning_rate": 9.63605977814564e-06,
"loss": 0.568,
"step": 500
},
{
"epoch": 0.1548771407297096,
"grad_norm": 0.12193777412176132,
"learning_rate": 9.59907279599747e-06,
"loss": 0.5661,
"step": 520
},
{
"epoch": 0.16083395383469842,
"grad_norm": 0.12415241450071335,
"learning_rate": 9.56037429756892e-06,
"loss": 0.5618,
"step": 540
},
{
"epoch": 0.16679076693968728,
"grad_norm": 0.12287413328886032,
"learning_rate": 9.51997868426502e-06,
"loss": 0.5731,
"step": 560
},
{
"epoch": 0.1727475800446761,
"grad_norm": 0.1268438696861267,
"learning_rate": 9.477900989061519e-06,
"loss": 0.5702,
"step": 580
},
{
"epoch": 0.17870439314966494,
"grad_norm": 0.12195920199155807,
"learning_rate": 9.43415687091047e-06,
"loss": 0.5649,
"step": 600
},
{
"epoch": 0.18466120625465376,
"grad_norm": 0.12017785012722015,
"learning_rate": 9.388762608912834e-06,
"loss": 0.561,
"step": 620
},
{
"epoch": 0.1906180193596426,
"grad_norm": 0.11770286411046982,
"learning_rate": 9.341735096260326e-06,
"loss": 0.566,
"step": 640
},
{
"epoch": 0.19657483246463142,
"grad_norm": 0.12057702243328094,
"learning_rate": 9.293091833948709e-06,
"loss": 0.5648,
"step": 660
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.12094268947839737,
"learning_rate": 9.242850924264918e-06,
"loss": 0.5584,
"step": 680
},
{
"epoch": 0.20848845867460908,
"grad_norm": 0.11510290950536728,
"learning_rate": 9.191031064050396e-06,
"loss": 0.5511,
"step": 700
},
{
"epoch": 0.2144452717795979,
"grad_norm": 0.11895924061536789,
"learning_rate": 9.137651537743193e-06,
"loss": 0.5548,
"step": 720
},
{
"epoch": 0.22040208488458674,
"grad_norm": 0.12275276333093643,
"learning_rate": 9.082732210201371e-06,
"loss": 0.556,
"step": 740
},
{
"epoch": 0.22635889798957556,
"grad_norm": 0.12446765601634979,
"learning_rate": 9.026293519310426e-06,
"loss": 0.5528,
"step": 760
},
{
"epoch": 0.23231571109456442,
"grad_norm": 0.12517477571964264,
"learning_rate": 8.968356468377454e-06,
"loss": 0.5448,
"step": 780
},
{
"epoch": 0.23827252419955325,
"grad_norm": 0.11592727899551392,
"learning_rate": 8.908942618314912e-06,
"loss": 0.549,
"step": 800
},
{
"epoch": 0.24422933730454208,
"grad_norm": 0.11930397152900696,
"learning_rate": 8.848074079616838e-06,
"loss": 0.5481,
"step": 820
},
{
"epoch": 0.2501861504095309,
"grad_norm": 0.11791594326496124,
"learning_rate": 8.785773504130587e-06,
"loss": 0.5449,
"step": 840
},
{
"epoch": 0.2561429635145197,
"grad_norm": 0.11881785094738007,
"learning_rate": 8.722064076627082e-06,
"loss": 0.5525,
"step": 860
},
{
"epoch": 0.26209977661950856,
"grad_norm": 0.11947067826986313,
"learning_rate": 8.65696950617274e-06,
"loss": 0.545,
"step": 880
},
{
"epoch": 0.2680565897244974,
"grad_norm": 0.1221371591091156,
"learning_rate": 8.590514017306298e-06,
"loss": 0.5532,
"step": 900
},
{
"epoch": 0.2740134028294862,
"grad_norm": 0.12159781903028488,
"learning_rate": 8.522722341023795e-06,
"loss": 0.5437,
"step": 920
},
{
"epoch": 0.2799702159344751,
"grad_norm": 0.11880378425121307,
"learning_rate": 8.453619705575096e-06,
"loss": 0.5398,
"step": 940
},
{
"epoch": 0.2859270290394639,
"grad_norm": 0.11256157606840134,
"learning_rate": 8.38323182707535e-06,
"loss": 0.5508,
"step": 960
},
{
"epoch": 0.29188384214445273,
"grad_norm": 0.11828071624040604,
"learning_rate": 8.31158489993489e-06,
"loss": 0.5387,
"step": 980
},
{
"epoch": 0.29784065524944153,
"grad_norm": 0.11970686912536621,
"learning_rate": 8.238705587111162e-06,
"loss": 0.548,
"step": 1000
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.1221531331539154,
"learning_rate": 8.16462101018626e-06,
"loss": 0.5416,
"step": 1020
},
{
"epoch": 0.3097542814594192,
"grad_norm": 0.11775178462266922,
"learning_rate": 8.089358739273806e-06,
"loss": 0.5382,
"step": 1040
},
{
"epoch": 0.31571109456440805,
"grad_norm": 0.11482196301221848,
"learning_rate": 8.0129467827589e-06,
"loss": 0.5379,
"step": 1060
},
{
"epoch": 0.32166790766939685,
"grad_norm": 0.11577887833118439,
"learning_rate": 7.935413576874966e-06,
"loss": 0.5434,
"step": 1080
},
{
"epoch": 0.3276247207743857,
"grad_norm": 0.12119460850954056,
"learning_rate": 7.856787975121407e-06,
"loss": 0.5442,
"step": 1100
},
{
"epoch": 0.33358153387937456,
"grad_norm": 0.11904456466436386,
"learning_rate": 7.777099237525929e-06,
"loss": 0.5323,
"step": 1120
},
{
"epoch": 0.33953834698436336,
"grad_norm": 0.11310964077711105,
"learning_rate": 7.696377019755624e-06,
"loss": 0.532,
"step": 1140
},
{
"epoch": 0.3454951600893522,
"grad_norm": 0.11461955308914185,
"learning_rate": 7.614651362080787e-06,
"loss": 0.5354,
"step": 1160
},
{
"epoch": 0.351451973194341,
"grad_norm": 0.11475788056850433,
"learning_rate": 7.531952678195621e-06,
"loss": 0.5393,
"step": 1180
},
{
"epoch": 0.3574087862993299,
"grad_norm": 0.11288796365261078,
"learning_rate": 7.4483117438999685e-06,
"loss": 0.5455,
"step": 1200
},
{
"epoch": 0.3633655994043187,
"grad_norm": 0.11906970292329788,
"learning_rate": 7.3637596856462945e-06,
"loss": 0.5444,
"step": 1220
},
{
"epoch": 0.36932241250930753,
"grad_norm": 0.1140429899096489,
"learning_rate": 7.278327968956159e-06,
"loss": 0.5323,
"step": 1240
},
{
"epoch": 0.37527922561429633,
"grad_norm": 0.11703943461179733,
"learning_rate": 7.192048386710537e-06,
"loss": 0.5358,
"step": 1260
},
{
"epoch": 0.3812360387192852,
"grad_norm": 0.11820199340581894,
"learning_rate": 7.10495304731828e-06,
"loss": 0.5421,
"step": 1280
},
{
"epoch": 0.387192851824274,
"grad_norm": 0.1111307144165039,
"learning_rate": 7.017074362767176e-06,
"loss": 0.5347,
"step": 1300
},
{
"epoch": 0.39314966492926284,
"grad_norm": 0.1102440133690834,
"learning_rate": 6.9284450365620385e-06,
"loss": 0.5318,
"step": 1320
},
{
"epoch": 0.3991064780342517,
"grad_norm": 0.10770157724618912,
"learning_rate": 6.83909805155429e-06,
"loss": 0.542,
"step": 1340
},
{
"epoch": 0.4050632911392405,
"grad_norm": 0.11749140918254852,
"learning_rate": 6.749066657667609e-06,
"loss": 0.5348,
"step": 1360
},
{
"epoch": 0.41102010424422936,
"grad_norm": 0.11329230666160583,
"learning_rate": 6.658384359524166e-06,
"loss": 0.5288,
"step": 1380
},
{
"epoch": 0.41697691734921816,
"grad_norm": 0.11443163454532623,
"learning_rate": 6.5670849039761e-06,
"loss": 0.5327,
"step": 1400
},
{
"epoch": 0.422933730454207,
"grad_norm": 0.1070597767829895,
"learning_rate": 6.475202267546829e-06,
"loss": 0.5328,
"step": 1420
},
{
"epoch": 0.4288905435591958,
"grad_norm": 0.11124689131975174,
"learning_rate": 6.382770643786907e-06,
"loss": 0.5263,
"step": 1440
},
{
"epoch": 0.43484735666418467,
"grad_norm": 0.11341281980276108,
"learning_rate": 6.289824430549092e-06,
"loss": 0.5348,
"step": 1460
},
{
"epoch": 0.44080416976917347,
"grad_norm": 0.11513620615005493,
"learning_rate": 6.196398217187418e-06,
"loss": 0.5323,
"step": 1480
},
{
"epoch": 0.4467609828741623,
"grad_norm": 0.11688452959060669,
"learning_rate": 6.102526771684971e-06,
"loss": 0.527,
"step": 1500
},
{
"epoch": 0.4527177959791511,
"grad_norm": 0.11436779797077179,
"learning_rate": 6.008245027715205e-06,
"loss": 0.5353,
"step": 1520
},
{
"epoch": 0.45867460908414,
"grad_norm": 0.1130824014544487,
"learning_rate": 5.9135880716415885e-06,
"loss": 0.5404,
"step": 1540
},
{
"epoch": 0.46463142218912884,
"grad_norm": 0.11337339878082275,
"learning_rate": 5.818591129460437e-06,
"loss": 0.5243,
"step": 1560
},
{
"epoch": 0.47058823529411764,
"grad_norm": 0.11147360503673553,
"learning_rate": 5.723289553691782e-06,
"loss": 0.5256,
"step": 1580
},
{
"epoch": 0.4765450483991065,
"grad_norm": 0.11559344828128815,
"learning_rate": 5.6277188102231396e-06,
"loss": 0.5352,
"step": 1600
},
{
"epoch": 0.4825018615040953,
"grad_norm": 0.11330032348632812,
"learning_rate": 5.531914465111109e-06,
"loss": 0.542,
"step": 1620
},
{
"epoch": 0.48845867460908415,
"grad_norm": 0.11136647313833237,
"learning_rate": 5.435912171345691e-06,
"loss": 0.5288,
"step": 1640
},
{
"epoch": 0.49441548771407295,
"grad_norm": 0.11813569068908691,
"learning_rate": 5.3397476555822365e-06,
"loss": 0.5309,
"step": 1660
},
{
"epoch": 0.5003723008190618,
"grad_norm": 0.11226009577512741,
"learning_rate": 5.243456704846011e-06,
"loss": 0.5262,
"step": 1680
},
{
"epoch": 0.5063291139240507,
"grad_norm": 0.12029188126325607,
"learning_rate": 5.147075153214256e-06,
"loss": 0.532,
"step": 1700
},
{
"epoch": 0.5122859270290394,
"grad_norm": 0.10719672590494156,
"learning_rate": 5.0506388684807805e-06,
"loss": 0.5293,
"step": 1720
},
{
"epoch": 0.5182427401340283,
"grad_norm": 0.10853017866611481,
"learning_rate": 4.954183738807972e-06,
"loss": 0.5349,
"step": 1740
},
{
"epoch": 0.5241995532390171,
"grad_norm": 0.1090475469827652,
"learning_rate": 4.857745659371249e-06,
"loss": 0.5262,
"step": 1760
},
{
"epoch": 0.530156366344006,
"grad_norm": 0.11181093752384186,
"learning_rate": 4.761360519000884e-06,
"loss": 0.5232,
"step": 1780
},
{
"epoch": 0.5361131794489948,
"grad_norm": 0.11282672733068466,
"learning_rate": 4.665064186826211e-06,
"loss": 0.524,
"step": 1800
},
{
"epoch": 0.5420699925539836,
"grad_norm": 0.1118716299533844,
"learning_rate": 4.568892498927148e-06,
"loss": 0.5268,
"step": 1820
},
{
"epoch": 0.5480268056589724,
"grad_norm": 0.11272594332695007,
"learning_rate": 4.47288124499802e-06,
"loss": 0.523,
"step": 1840
},
{
"epoch": 0.5539836187639613,
"grad_norm": 0.1089363694190979,
"learning_rate": 4.3770661550286485e-06,
"loss": 0.5205,
"step": 1860
},
{
"epoch": 0.5599404318689502,
"grad_norm": 0.11689860373735428,
"learning_rate": 4.281482886007665e-06,
"loss": 0.529,
"step": 1880
},
{
"epoch": 0.5658972449739389,
"grad_norm": 0.10627499967813492,
"learning_rate": 4.186167008652975e-06,
"loss": 0.5249,
"step": 1900
},
{
"epoch": 0.5718540580789278,
"grad_norm": 0.11572492122650146,
"learning_rate": 4.091153994174343e-06,
"loss": 0.5276,
"step": 1920
},
{
"epoch": 0.5778108711839166,
"grad_norm": 0.11439114809036255,
"learning_rate": 3.996479201073005e-06,
"loss": 0.5192,
"step": 1940
},
{
"epoch": 0.5837676842889055,
"grad_norm": 0.11502502113580704,
"learning_rate": 3.902177861983213e-06,
"loss": 0.5223,
"step": 1960
},
{
"epoch": 0.5897244973938943,
"grad_norm": 0.11334360390901566,
"learning_rate": 3.8082850705606387e-06,
"loss": 0.5263,
"step": 1980
},
{
"epoch": 0.5956813104988831,
"grad_norm": 0.11678946018218994,
"learning_rate": 3.7148357684224887e-06,
"loss": 0.5306,
"step": 2000
},
{
"epoch": 0.6016381236038719,
"grad_norm": 0.10717915743589401,
"learning_rate": 3.6218647321441934e-06,
"loss": 0.5224,
"step": 2020
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.1125260517001152,
"learning_rate": 3.529406560317521e-06,
"loss": 0.5256,
"step": 2040
},
{
"epoch": 0.6135517498138496,
"grad_norm": 0.10978550463914871,
"learning_rate": 3.4374956606749337e-06,
"loss": 0.5213,
"step": 2060
},
{
"epoch": 0.6195085629188384,
"grad_norm": 0.10976285487413406,
"learning_rate": 3.3461662372849635e-06,
"loss": 0.5166,
"step": 2080
},
{
"epoch": 0.6254653760238272,
"grad_norm": 0.1127789244055748,
"learning_rate": 3.25545227782337e-06,
"loss": 0.5283,
"step": 2100
},
{
"epoch": 0.6314221891288161,
"grad_norm": 0.1105373352766037,
"learning_rate": 3.1653875409248557e-06,
"loss": 0.5202,
"step": 2120
},
{
"epoch": 0.637379002233805,
"grad_norm": 0.11276114732027054,
"learning_rate": 3.07600554361999e-06,
"loss": 0.5206,
"step": 2140
},
{
"epoch": 0.6433358153387937,
"grad_norm": 0.11418343335390091,
"learning_rate": 2.987339548862054e-06,
"loss": 0.5257,
"step": 2160
},
{
"epoch": 0.6492926284437825,
"grad_norm": 0.11187276989221573,
"learning_rate": 2.8994225531484456e-06,
"loss": 0.5242,
"step": 2180
},
{
"epoch": 0.6552494415487714,
"grad_norm": 0.11579373478889465,
"learning_rate": 2.812287274241238e-06,
"loss": 0.52,
"step": 2200
},
{
"epoch": 0.6612062546537603,
"grad_norm": 0.11193983256816864,
"learning_rate": 2.7259661389914597e-06,
"loss": 0.5239,
"step": 2220
},
{
"epoch": 0.6671630677587491,
"grad_norm": 0.11156244575977325,
"learning_rate": 2.6404912712716585e-06,
"loss": 0.5242,
"step": 2240
},
{
"epoch": 0.6731198808637379,
"grad_norm": 0.11163734644651413,
"learning_rate": 2.5558944800211892e-06,
"loss": 0.5245,
"step": 2260
},
{
"epoch": 0.6790766939687267,
"grad_norm": 0.11284824460744858,
"learning_rate": 2.472207247408727e-06,
"loss": 0.5265,
"step": 2280
},
{
"epoch": 0.6850335070737156,
"grad_norm": 0.10937218368053436,
"learning_rate": 2.389460717116358e-06,
"loss": 0.5194,
"step": 2300
},
{
"epoch": 0.6909903201787044,
"grad_norm": 0.11130936443805695,
"learning_rate": 2.30768568274967e-06,
"loss": 0.5205,
"step": 2320
},
{
"epoch": 0.6969471332836932,
"grad_norm": 0.11463384330272675,
"learning_rate": 2.2269125763780974e-06,
"loss": 0.5146,
"step": 2340
},
{
"epoch": 0.702903946388682,
"grad_norm": 0.10185515135526657,
"learning_rate": 2.147171457209807e-06,
"loss": 0.5166,
"step": 2360
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.11237939447164536,
"learning_rate": 2.0684920004053666e-06,
"loss": 0.5159,
"step": 2380
},
{
"epoch": 0.7148175725986597,
"grad_norm": 0.11210869252681732,
"learning_rate": 1.990903486034308e-06,
"loss": 0.5228,
"step": 2400
},
{
"epoch": 0.7207743857036486,
"grad_norm": 0.11043533682823181,
"learning_rate": 1.914434788178737e-06,
"loss": 0.5216,
"step": 2420
},
{
"epoch": 0.7267311988086373,
"grad_norm": 0.11434119939804077,
"learning_rate": 1.8391143641880188e-06,
"loss": 0.5131,
"step": 2440
},
{
"epoch": 0.7326880119136262,
"grad_norm": 0.10699808597564697,
"learning_rate": 1.7649702440885597e-06,
"loss": 0.5229,
"step": 2460
},
{
"epoch": 0.7386448250186151,
"grad_norm": 0.11019433289766312,
"learning_rate": 1.692030020152598e-06,
"loss": 0.5234,
"step": 2480
},
{
"epoch": 0.7446016381236039,
"grad_norm": 0.10579477250576019,
"learning_rate": 1.6203208366299216e-06,
"loss": 0.5267,
"step": 2500
},
{
"epoch": 0.7505584512285927,
"grad_norm": 0.10574056208133698,
"learning_rate": 1.5498693796462994e-06,
"loss": 0.5222,
"step": 2520
},
{
"epoch": 0.7565152643335815,
"grad_norm": 0.11808321624994278,
"learning_rate": 1.4807018672724066e-06,
"loss": 0.5158,
"step": 2540
},
{
"epoch": 0.7624720774385704,
"grad_norm": 0.11037767678499222,
"learning_rate": 1.4128440397669301e-06,
"loss": 0.5168,
"step": 2560
},
{
"epoch": 0.7684288905435592,
"grad_norm": 0.1146508976817131,
"learning_rate": 1.3463211499974993e-06,
"loss": 0.5185,
"step": 2580
},
{
"epoch": 0.774385703648548,
"grad_norm": 0.11421221494674683,
"learning_rate": 1.2811579540429836e-06,
"loss": 0.5218,
"step": 2600
},
{
"epoch": 0.7803425167535368,
"grad_norm": 0.11680857837200165,
"learning_rate": 1.2173787019806688e-06,
"loss": 0.5172,
"step": 2620
},
{
"epoch": 0.7862993298585257,
"grad_norm": 0.11102883517742157,
"learning_rate": 1.1550071288617486e-06,
"loss": 0.5202,
"step": 2640
},
{
"epoch": 0.7922561429635145,
"grad_norm": 0.1098845824599266,
"learning_rate": 1.0940664458784733e-06,
"loss": 0.5168,
"step": 2660
},
{
"epoch": 0.7982129560685034,
"grad_norm": 0.11351136863231659,
"learning_rate": 1.0345793317262387e-06,
"loss": 0.5163,
"step": 2680
},
{
"epoch": 0.8041697691734921,
"grad_norm": 0.11287049204111099,
"learning_rate": 9.765679241638649e-07,
"loss": 0.5254,
"step": 2700
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.11457482725381851,
"learning_rate": 9.200538117751595e-07,
"loss": 0.5272,
"step": 2720
},
{
"epoch": 0.8160833953834699,
"grad_norm": 0.11306252330541611,
"learning_rate": 8.650580259348595e-07,
"loss": 0.5261,
"step": 2740
},
{
"epoch": 0.8220402084884587,
"grad_norm": 0.1157909482717514,
"learning_rate": 8.116010329819418e-07,
"loss": 0.523,
"step": 2760
},
{
"epoch": 0.8279970215934475,
"grad_norm": 0.1141168624162674,
"learning_rate": 7.597027266031992e-07,
"loss": 0.5178,
"step": 2780
},
{
"epoch": 0.8339538346984363,
"grad_norm": 0.10697868466377258,
"learning_rate": 7.093824204299299e-07,
"loss": 0.5182,
"step": 2800
},
{
"epoch": 0.8399106478034252,
"grad_norm": 0.11109849065542221,
"learning_rate": 6.60658840850486e-07,
"loss": 0.5324,
"step": 2820
},
{
"epoch": 0.845867460908414,
"grad_norm": 0.10724302381277084,
"learning_rate": 6.135501200413724e-07,
"loss": 0.5168,
"step": 2840
},
{
"epoch": 0.8518242740134029,
"grad_norm": 0.11035523563623428,
"learning_rate": 5.680737892194654e-07,
"loss": 0.5226,
"step": 2860
},
{
"epoch": 0.8577810871183916,
"grad_norm": 0.11459671705961227,
"learning_rate": 5.242467721178774e-07,
"loss": 0.5215,
"step": 2880
},
{
"epoch": 0.8637379002233805,
"grad_norm": 0.11318153887987137,
"learning_rate": 4.82085378687901e-07,
"loss": 0.5155,
"step": 2900
},
{
"epoch": 0.8696947133283693,
"grad_norm": 0.11405057460069656,
"learning_rate": 4.416052990293596e-07,
"loss": 0.5164,
"step": 2920
},
{
"epoch": 0.8756515264333582,
"grad_norm": 0.1100643128156662,
"learning_rate": 4.028215975516314e-07,
"loss": 0.5214,
"step": 2940
},
{
"epoch": 0.8816083395383469,
"grad_norm": 0.11574219167232513,
"learning_rate": 3.657487073675281e-07,
"loss": 0.5303,
"step": 2960
},
{
"epoch": 0.8875651526433358,
"grad_norm": 0.10859781503677368,
"learning_rate": 3.3040042492210103e-07,
"loss": 0.515,
"step": 2980
},
{
"epoch": 0.8935219657483247,
"grad_norm": 0.12005914002656937,
"learning_rate": 2.9678990485837553e-07,
"loss": 0.5297,
"step": 3000
},
{
"epoch": 0.8994787788533135,
"grad_norm": 0.10555088520050049,
"learning_rate": 2.6492965512193934e-07,
"loss": 0.5236,
"step": 3020
},
{
"epoch": 0.9054355919583023,
"grad_norm": 0.10925314575433731,
"learning_rate": 2.348315323061845e-07,
"loss": 0.5343,
"step": 3040
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.10835836082696915,
"learning_rate": 2.0650673723995097e-07,
"loss": 0.5119,
"step": 3060
},
{
"epoch": 0.91734921816828,
"grad_norm": 0.11512824147939682,
"learning_rate": 1.7996581081920483e-07,
"loss": 0.5165,
"step": 3080
},
{
"epoch": 0.9233060312732688,
"grad_norm": 0.11410706490278244,
"learning_rate": 1.5521863008431725e-07,
"loss": 0.522,
"step": 3100
},
{
"epoch": 0.9292628443782577,
"grad_norm": 0.11093121767044067,
"learning_rate": 1.3227440454437745e-07,
"loss": 0.5179,
"step": 3120
},
{
"epoch": 0.9352196574832464,
"grad_norm": 0.11221588402986526,
"learning_rate": 1.1114167274993537e-07,
"loss": 0.5326,
"step": 3140
},
{
"epoch": 0.9411764705882353,
"grad_norm": 0.11275220662355423,
"learning_rate": 9.1828299115434e-08,
"loss": 0.5227,
"step": 3160
},
{
"epoch": 0.9471332836932241,
"grad_norm": 0.1078876256942749,
"learning_rate": 7.434147099251165e-08,
"loss": 0.5191,
"step": 3180
},
{
"epoch": 0.953090096798213,
"grad_norm": 0.1099793091416359,
"learning_rate": 5.86876959952809e-08,
"loss": 0.517,
"step": 3200
},
{
"epoch": 0.9590469099032017,
"grad_norm": 0.11506114900112152,
"learning_rate": 4.4872799578554724e-08,
"loss": 0.5202,
"step": 3220
},
{
"epoch": 0.9650037230081906,
"grad_norm": 0.10891096293926239,
"learning_rate": 3.290192286994287e-08,
"loss": 0.5236,
"step": 3240
},
{
"epoch": 0.9709605361131795,
"grad_norm": 0.10671822726726532,
"learning_rate": 2.2779520756611163e-08,
"loss": 0.5124,
"step": 3260
},
{
"epoch": 0.9769173492181683,
"grad_norm": 0.11090593785047531,
"learning_rate": 1.450936022742544e-08,
"loss": 0.5284,
"step": 3280
},
{
"epoch": 0.9828741623231572,
"grad_norm": 0.11483392119407654,
"learning_rate": 8.094518971088484e-09,
"loss": 0.5236,
"step": 3300
},
{
"epoch": 0.9888309754281459,
"grad_norm": 0.11123649030923843,
"learning_rate": 3.537384230797303e-09,
"loss": 0.5184,
"step": 3320
},
{
"epoch": 0.9947877885331348,
"grad_norm": 0.11426061391830444,
"learning_rate": 8.396519158465621e-10,
"loss": 0.5169,
"step": 3340
},
{
"epoch": 1.0,
"step": 3358,
"total_flos": 7.06306771390772e+18,
"train_loss": 0.5600698575013588,
"train_runtime": 4520.9828,
"train_samples_per_second": 190.112,
"train_steps_per_second": 0.743
}
],
"logging_steps": 20,
"max_steps": 3358,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 7.06306771390772e+18,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}