Files
math-1b-sft-numinamath-bs51…/trainer_state.json

1213 lines
29 KiB
JSON
Raw Permalink Normal View History

{
"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": 2.119563341140747,
"learning_rate": 1.8811881188118813e-06,
"loss": 1.3592,
"step": 20
},
{
"epoch": 0.011913626209977662,
"grad_norm": 0.4560997486114502,
"learning_rate": 3.861386138613862e-06,
"loss": 1.0573,
"step": 40
},
{
"epoch": 0.017870439314966492,
"grad_norm": 0.23475755751132965,
"learning_rate": 5.841584158415842e-06,
"loss": 0.8295,
"step": 60
},
{
"epoch": 0.023827252419955324,
"grad_norm": 0.18663567304611206,
"learning_rate": 7.821782178217822e-06,
"loss": 0.7659,
"step": 80
},
{
"epoch": 0.029784065524944156,
"grad_norm": 0.18412990868091583,
"learning_rate": 9.801980198019802e-06,
"loss": 0.7285,
"step": 100
},
{
"epoch": 0.035740878629932984,
"grad_norm": 0.1868126392364502,
"learning_rate": 9.99924640442892e-06,
"loss": 0.7129,
"step": 120
},
{
"epoch": 0.04169769173492182,
"grad_norm": 0.1749425232410431,
"learning_rate": 9.99664167434275e-06,
"loss": 0.7024,
"step": 140
},
{
"epoch": 0.04765450483991065,
"grad_norm": 0.1737823337316513,
"learning_rate": 9.99217747519571e-06,
"loss": 0.6906,
"step": 160
},
{
"epoch": 0.05361131794489948,
"grad_norm": 0.17928074300289154,
"learning_rate": 9.98585546831169e-06,
"loss": 0.6919,
"step": 180
},
{
"epoch": 0.05956813104988831,
"grad_norm": 0.17608019709587097,
"learning_rate": 9.977678006386153e-06,
"loss": 0.6717,
"step": 200
},
{
"epoch": 0.06552494415487714,
"grad_norm": 0.17232322692871094,
"learning_rate": 9.967648132610596e-06,
"loss": 0.6739,
"step": 220
},
{
"epoch": 0.07148175725986597,
"grad_norm": 0.17418703436851501,
"learning_rate": 9.95576957954004e-06,
"loss": 0.6725,
"step": 240
},
{
"epoch": 0.0774385703648548,
"grad_norm": 0.17881107330322266,
"learning_rate": 9.942046767703982e-06,
"loss": 0.6693,
"step": 260
},
{
"epoch": 0.08339538346984364,
"grad_norm": 0.16734157502651215,
"learning_rate": 9.926484803961332e-06,
"loss": 0.6639,
"step": 280
},
{
"epoch": 0.08935219657483247,
"grad_norm": 0.1772398054599762,
"learning_rate": 9.909089479599916e-06,
"loss": 0.666,
"step": 300
},
{
"epoch": 0.0953090096798213,
"grad_norm": 0.1695602387189865,
"learning_rate": 9.88986726818129e-06,
"loss": 0.6663,
"step": 320
},
{
"epoch": 0.10126582278481013,
"grad_norm": 0.17020754516124725,
"learning_rate": 9.868825323131646e-06,
"loss": 0.653,
"step": 340
},
{
"epoch": 0.10722263588979895,
"grad_norm": 0.17200911045074463,
"learning_rate": 9.8459714750797e-06,
"loss": 0.6531,
"step": 360
},
{
"epoch": 0.11317944899478778,
"grad_norm": 0.17503520846366882,
"learning_rate": 9.821314228942587e-06,
"loss": 0.6632,
"step": 380
},
{
"epoch": 0.11913626209977662,
"grad_norm": 0.17521335184574127,
"learning_rate": 9.794862760760798e-06,
"loss": 0.652,
"step": 400
},
{
"epoch": 0.12509307520476545,
"grad_norm": 0.17581316828727722,
"learning_rate": 9.76662691428337e-06,
"loss": 0.6572,
"step": 420
},
{
"epoch": 0.13104988830975428,
"grad_norm": 0.17654919624328613,
"learning_rate": 9.736617197304615e-06,
"loss": 0.6454,
"step": 440
},
{
"epoch": 0.1370067014147431,
"grad_norm": 0.1758466362953186,
"learning_rate": 9.704844777753686e-06,
"loss": 0.6408,
"step": 460
},
{
"epoch": 0.14296351451973194,
"grad_norm": 0.1721346229314804,
"learning_rate": 9.671321479538522e-06,
"loss": 0.6418,
"step": 480
},
{
"epoch": 0.14892032762472077,
"grad_norm": 0.17906631529331207,
"learning_rate": 9.63605977814564e-06,
"loss": 0.6396,
"step": 500
},
{
"epoch": 0.1548771407297096,
"grad_norm": 0.18055592477321625,
"learning_rate": 9.59907279599747e-06,
"loss": 0.6382,
"step": 520
},
{
"epoch": 0.16083395383469842,
"grad_norm": 0.18354107439517975,
"learning_rate": 9.56037429756892e-06,
"loss": 0.6325,
"step": 540
},
{
"epoch": 0.16679076693968728,
"grad_norm": 0.17623865604400635,
"learning_rate": 9.51997868426502e-06,
"loss": 0.645,
"step": 560
},
{
"epoch": 0.1727475800446761,
"grad_norm": 0.1768830269575119,
"learning_rate": 9.477900989061519e-06,
"loss": 0.6403,
"step": 580
},
{
"epoch": 0.17870439314966494,
"grad_norm": 0.1716880053281784,
"learning_rate": 9.43415687091047e-06,
"loss": 0.6362,
"step": 600
},
{
"epoch": 0.18466120625465376,
"grad_norm": 0.17497296631336212,
"learning_rate": 9.388762608912834e-06,
"loss": 0.6321,
"step": 620
},
{
"epoch": 0.1906180193596426,
"grad_norm": 0.1781647503376007,
"learning_rate": 9.341735096260326e-06,
"loss": 0.6386,
"step": 640
},
{
"epoch": 0.19657483246463142,
"grad_norm": 0.1796586662530899,
"learning_rate": 9.293091833948709e-06,
"loss": 0.637,
"step": 660
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.17507584393024445,
"learning_rate": 9.242850924264918e-06,
"loss": 0.6299,
"step": 680
},
{
"epoch": 0.20848845867460908,
"grad_norm": 0.16802071034908295,
"learning_rate": 9.191031064050396e-06,
"loss": 0.6223,
"step": 700
},
{
"epoch": 0.2144452717795979,
"grad_norm": 0.1792314648628235,
"learning_rate": 9.137651537743193e-06,
"loss": 0.6277,
"step": 720
},
{
"epoch": 0.22040208488458674,
"grad_norm": 0.17301903665065765,
"learning_rate": 9.082732210201371e-06,
"loss": 0.6275,
"step": 740
},
{
"epoch": 0.22635889798957556,
"grad_norm": 0.17726927995681763,
"learning_rate": 9.026293519310426e-06,
"loss": 0.6245,
"step": 760
},
{
"epoch": 0.23231571109456442,
"grad_norm": 0.17860877513885498,
"learning_rate": 8.968356468377454e-06,
"loss": 0.6173,
"step": 780
},
{
"epoch": 0.23827252419955325,
"grad_norm": 0.16912327706813812,
"learning_rate": 8.908942618314912e-06,
"loss": 0.6191,
"step": 800
},
{
"epoch": 0.24422933730454208,
"grad_norm": 0.17608532309532166,
"learning_rate": 8.848074079616838e-06,
"loss": 0.619,
"step": 820
},
{
"epoch": 0.2501861504095309,
"grad_norm": 0.16739599406719208,
"learning_rate": 8.785773504130587e-06,
"loss": 0.6167,
"step": 840
},
{
"epoch": 0.2561429635145197,
"grad_norm": 0.17052432894706726,
"learning_rate": 8.722064076627082e-06,
"loss": 0.6248,
"step": 860
},
{
"epoch": 0.26209977661950856,
"grad_norm": 0.17976507544517517,
"learning_rate": 8.65696950617274e-06,
"loss": 0.6176,
"step": 880
},
{
"epoch": 0.2680565897244974,
"grad_norm": 0.17950667440891266,
"learning_rate": 8.590514017306298e-06,
"loss": 0.6251,
"step": 900
},
{
"epoch": 0.2740134028294862,
"grad_norm": 0.1708395630121231,
"learning_rate": 8.522722341023795e-06,
"loss": 0.6138,
"step": 920
},
{
"epoch": 0.2799702159344751,
"grad_norm": 0.16705350577831268,
"learning_rate": 8.453619705575096e-06,
"loss": 0.6099,
"step": 940
},
{
"epoch": 0.2859270290394639,
"grad_norm": 0.16709111630916595,
"learning_rate": 8.38323182707535e-06,
"loss": 0.6222,
"step": 960
},
{
"epoch": 0.29188384214445273,
"grad_norm": 0.16841235756874084,
"learning_rate": 8.31158489993489e-06,
"loss": 0.6091,
"step": 980
},
{
"epoch": 0.29784065524944153,
"grad_norm": 0.17064417898654938,
"learning_rate": 8.238705587111162e-06,
"loss": 0.6187,
"step": 1000
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.18553495407104492,
"learning_rate": 8.16462101018626e-06,
"loss": 0.6127,
"step": 1020
},
{
"epoch": 0.3097542814594192,
"grad_norm": 0.173476442694664,
"learning_rate": 8.089358739273806e-06,
"loss": 0.6092,
"step": 1040
},
{
"epoch": 0.31571109456440805,
"grad_norm": 0.1615096926689148,
"learning_rate": 8.0129467827589e-06,
"loss": 0.6083,
"step": 1060
},
{
"epoch": 0.32166790766939685,
"grad_norm": 0.1721673160791397,
"learning_rate": 7.935413576874966e-06,
"loss": 0.6141,
"step": 1080
},
{
"epoch": 0.3276247207743857,
"grad_norm": 0.17340931296348572,
"learning_rate": 7.856787975121407e-06,
"loss": 0.6145,
"step": 1100
},
{
"epoch": 0.33358153387937456,
"grad_norm": 0.17477653920650482,
"learning_rate": 7.777099237525929e-06,
"loss": 0.602,
"step": 1120
},
{
"epoch": 0.33953834698436336,
"grad_norm": 0.17575432360172272,
"learning_rate": 7.696377019755624e-06,
"loss": 0.6024,
"step": 1140
},
{
"epoch": 0.3454951600893522,
"grad_norm": 0.16720600426197052,
"learning_rate": 7.614651362080787e-06,
"loss": 0.6061,
"step": 1160
},
{
"epoch": 0.351451973194341,
"grad_norm": 0.1643887758255005,
"learning_rate": 7.531952678195621e-06,
"loss": 0.6101,
"step": 1180
},
{
"epoch": 0.3574087862993299,
"grad_norm": 0.16761791706085205,
"learning_rate": 7.4483117438999685e-06,
"loss": 0.617,
"step": 1200
},
{
"epoch": 0.3633655994043187,
"grad_norm": 0.1738751232624054,
"learning_rate": 7.3637596856462945e-06,
"loss": 0.615,
"step": 1220
},
{
"epoch": 0.36932241250930753,
"grad_norm": 0.16891705989837646,
"learning_rate": 7.278327968956159e-06,
"loss": 0.602,
"step": 1240
},
{
"epoch": 0.37527922561429633,
"grad_norm": 0.17447391152381897,
"learning_rate": 7.192048386710537e-06,
"loss": 0.6069,
"step": 1260
},
{
"epoch": 0.3812360387192852,
"grad_norm": 0.1732683926820755,
"learning_rate": 7.10495304731828e-06,
"loss": 0.6134,
"step": 1280
},
{
"epoch": 0.387192851824274,
"grad_norm": 0.1645285189151764,
"learning_rate": 7.017074362767176e-06,
"loss": 0.6049,
"step": 1300
},
{
"epoch": 0.39314966492926284,
"grad_norm": 0.16298213601112366,
"learning_rate": 6.9284450365620385e-06,
"loss": 0.6024,
"step": 1320
},
{
"epoch": 0.3991064780342517,
"grad_norm": 0.16107381880283356,
"learning_rate": 6.83909805155429e-06,
"loss": 0.6133,
"step": 1340
},
{
"epoch": 0.4050632911392405,
"grad_norm": 0.17286071181297302,
"learning_rate": 6.749066657667609e-06,
"loss": 0.6051,
"step": 1360
},
{
"epoch": 0.41102010424422936,
"grad_norm": 0.16563032567501068,
"learning_rate": 6.658384359524166e-06,
"loss": 0.5988,
"step": 1380
},
{
"epoch": 0.41697691734921816,
"grad_norm": 0.1716582179069519,
"learning_rate": 6.5670849039761e-06,
"loss": 0.6025,
"step": 1400
},
{
"epoch": 0.422933730454207,
"grad_norm": 0.16528673470020294,
"learning_rate": 6.475202267546829e-06,
"loss": 0.6027,
"step": 1420
},
{
"epoch": 0.4288905435591958,
"grad_norm": 0.167840376496315,
"learning_rate": 6.382770643786907e-06,
"loss": 0.596,
"step": 1440
},
{
"epoch": 0.43484735666418467,
"grad_norm": 0.16823585331439972,
"learning_rate": 6.289824430549092e-06,
"loss": 0.6054,
"step": 1460
},
{
"epoch": 0.44080416976917347,
"grad_norm": 0.1716415137052536,
"learning_rate": 6.196398217187418e-06,
"loss": 0.6024,
"step": 1480
},
{
"epoch": 0.4467609828741623,
"grad_norm": 0.1692785918712616,
"learning_rate": 6.102526771684971e-06,
"loss": 0.5967,
"step": 1500
},
{
"epoch": 0.4527177959791511,
"grad_norm": 0.1642398089170456,
"learning_rate": 6.008245027715205e-06,
"loss": 0.6049,
"step": 1520
},
{
"epoch": 0.45867460908414,
"grad_norm": 0.16624784469604492,
"learning_rate": 5.9135880716415885e-06,
"loss": 0.6101,
"step": 1540
},
{
"epoch": 0.46463142218912884,
"grad_norm": 0.16547992825508118,
"learning_rate": 5.818591129460437e-06,
"loss": 0.5935,
"step": 1560
},
{
"epoch": 0.47058823529411764,
"grad_norm": 0.16697312891483307,
"learning_rate": 5.723289553691782e-06,
"loss": 0.5953,
"step": 1580
},
{
"epoch": 0.4765450483991065,
"grad_norm": 0.17255817353725433,
"learning_rate": 5.6277188102231396e-06,
"loss": 0.605,
"step": 1600
},
{
"epoch": 0.4825018615040953,
"grad_norm": 0.1701556146144867,
"learning_rate": 5.531914465111109e-06,
"loss": 0.6118,
"step": 1620
},
{
"epoch": 0.48845867460908415,
"grad_norm": 0.16841056942939758,
"learning_rate": 5.435912171345691e-06,
"loss": 0.598,
"step": 1640
},
{
"epoch": 0.49441548771407295,
"grad_norm": 0.1725454479455948,
"learning_rate": 5.3397476555822365e-06,
"loss": 0.6006,
"step": 1660
},
{
"epoch": 0.5003723008190618,
"grad_norm": 0.16914981603622437,
"learning_rate": 5.243456704846011e-06,
"loss": 0.5962,
"step": 1680
},
{
"epoch": 0.5063291139240507,
"grad_norm": 0.17998944222927094,
"learning_rate": 5.147075153214256e-06,
"loss": 0.6018,
"step": 1700
},
{
"epoch": 0.5122859270290394,
"grad_norm": 0.16358016431331635,
"learning_rate": 5.0506388684807805e-06,
"loss": 0.5993,
"step": 1720
},
{
"epoch": 0.5182427401340283,
"grad_norm": 0.16756755113601685,
"learning_rate": 4.954183738807972e-06,
"loss": 0.6042,
"step": 1740
},
{
"epoch": 0.5241995532390171,
"grad_norm": 0.1693088263273239,
"learning_rate": 4.857745659371249e-06,
"loss": 0.5958,
"step": 1760
},
{
"epoch": 0.530156366344006,
"grad_norm": 0.1687222272157669,
"learning_rate": 4.761360519000884e-06,
"loss": 0.5927,
"step": 1780
},
{
"epoch": 0.5361131794489948,
"grad_norm": 0.16474668681621552,
"learning_rate": 4.665064186826211e-06,
"loss": 0.5931,
"step": 1800
},
{
"epoch": 0.5420699925539836,
"grad_norm": 0.16657961905002594,
"learning_rate": 4.568892498927148e-06,
"loss": 0.5963,
"step": 1820
},
{
"epoch": 0.5480268056589724,
"grad_norm": 0.16624104976654053,
"learning_rate": 4.47288124499802e-06,
"loss": 0.591,
"step": 1840
},
{
"epoch": 0.5539836187639613,
"grad_norm": 0.16329768300056458,
"learning_rate": 4.3770661550286485e-06,
"loss": 0.5891,
"step": 1860
},
{
"epoch": 0.5599404318689502,
"grad_norm": 0.1724916696548462,
"learning_rate": 4.281482886007665e-06,
"loss": 0.5981,
"step": 1880
},
{
"epoch": 0.5658972449739389,
"grad_norm": 0.16009803116321564,
"learning_rate": 4.186167008652975e-06,
"loss": 0.5926,
"step": 1900
},
{
"epoch": 0.5718540580789278,
"grad_norm": 0.1676182895898819,
"learning_rate": 4.091153994174343e-06,
"loss": 0.596,
"step": 1920
},
{
"epoch": 0.5778108711839166,
"grad_norm": 0.17324478924274445,
"learning_rate": 3.996479201073005e-06,
"loss": 0.5881,
"step": 1940
},
{
"epoch": 0.5837676842889055,
"grad_norm": 0.17399553954601288,
"learning_rate": 3.902177861983213e-06,
"loss": 0.5907,
"step": 1960
},
{
"epoch": 0.5897244973938943,
"grad_norm": 0.17045123875141144,
"learning_rate": 3.8082850705606387e-06,
"loss": 0.5956,
"step": 1980
},
{
"epoch": 0.5956813104988831,
"grad_norm": 0.1752062439918518,
"learning_rate": 3.7148357684224887e-06,
"loss": 0.6,
"step": 2000
},
{
"epoch": 0.6016381236038719,
"grad_norm": 0.16427180171012878,
"learning_rate": 3.6218647321441934e-06,
"loss": 0.5914,
"step": 2020
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.1690833568572998,
"learning_rate": 3.529406560317521e-06,
"loss": 0.5937,
"step": 2040
},
{
"epoch": 0.6135517498138496,
"grad_norm": 0.16374030709266663,
"learning_rate": 3.4374956606749337e-06,
"loss": 0.5899,
"step": 2060
},
{
"epoch": 0.6195085629188384,
"grad_norm": 0.16435906291007996,
"learning_rate": 3.3461662372849635e-06,
"loss": 0.585,
"step": 2080
},
{
"epoch": 0.6254653760238272,
"grad_norm": 0.17175805568695068,
"learning_rate": 3.25545227782337e-06,
"loss": 0.597,
"step": 2100
},
{
"epoch": 0.6314221891288161,
"grad_norm": 0.1640237718820572,
"learning_rate": 3.1653875409248557e-06,
"loss": 0.5886,
"step": 2120
},
{
"epoch": 0.637379002233805,
"grad_norm": 0.17509537935256958,
"learning_rate": 3.07600554361999e-06,
"loss": 0.589,
"step": 2140
},
{
"epoch": 0.6433358153387937,
"grad_norm": 0.1702396720647812,
"learning_rate": 2.987339548862054e-06,
"loss": 0.5937,
"step": 2160
},
{
"epoch": 0.6492926284437825,
"grad_norm": 0.16526927053928375,
"learning_rate": 2.8994225531484456e-06,
"loss": 0.5929,
"step": 2180
},
{
"epoch": 0.6552494415487714,
"grad_norm": 0.1713491976261139,
"learning_rate": 2.812287274241238e-06,
"loss": 0.588,
"step": 2200
},
{
"epoch": 0.6612062546537603,
"grad_norm": 0.16687653958797455,
"learning_rate": 2.7259661389914597e-06,
"loss": 0.5917,
"step": 2220
},
{
"epoch": 0.6671630677587491,
"grad_norm": 0.16643133759498596,
"learning_rate": 2.6404912712716585e-06,
"loss": 0.5922,
"step": 2240
},
{
"epoch": 0.6731198808637379,
"grad_norm": 0.17374736070632935,
"learning_rate": 2.5558944800211892e-06,
"loss": 0.5918,
"step": 2260
},
{
"epoch": 0.6790766939687267,
"grad_norm": 0.1663503795862198,
"learning_rate": 2.472207247408727e-06,
"loss": 0.5956,
"step": 2280
},
{
"epoch": 0.6850335070737156,
"grad_norm": 0.16165192425251007,
"learning_rate": 2.389460717116358e-06,
"loss": 0.5879,
"step": 2300
},
{
"epoch": 0.6909903201787044,
"grad_norm": 0.16894033551216125,
"learning_rate": 2.30768568274967e-06,
"loss": 0.5875,
"step": 2320
},
{
"epoch": 0.6969471332836932,
"grad_norm": 0.1717192679643631,
"learning_rate": 2.2269125763780974e-06,
"loss": 0.5828,
"step": 2340
},
{
"epoch": 0.702903946388682,
"grad_norm": 0.16003255546092987,
"learning_rate": 2.147171457209807e-06,
"loss": 0.5838,
"step": 2360
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.16806617379188538,
"learning_rate": 2.0684920004053666e-06,
"loss": 0.5831,
"step": 2380
},
{
"epoch": 0.7148175725986597,
"grad_norm": 0.16845034062862396,
"learning_rate": 1.990903486034308e-06,
"loss": 0.5904,
"step": 2400
},
{
"epoch": 0.7207743857036486,
"grad_norm": 0.16814328730106354,
"learning_rate": 1.914434788178737e-06,
"loss": 0.589,
"step": 2420
},
{
"epoch": 0.7267311988086373,
"grad_norm": 0.16969594359397888,
"learning_rate": 1.8391143641880188e-06,
"loss": 0.5804,
"step": 2440
},
{
"epoch": 0.7326880119136262,
"grad_norm": 0.1628410369157791,
"learning_rate": 1.7649702440885597e-06,
"loss": 0.5899,
"step": 2460
},
{
"epoch": 0.7386448250186151,
"grad_norm": 0.1683746576309204,
"learning_rate": 1.692030020152598e-06,
"loss": 0.592,
"step": 2480
},
{
"epoch": 0.7446016381236039,
"grad_norm": 0.1615367978811264,
"learning_rate": 1.6203208366299216e-06,
"loss": 0.5954,
"step": 2500
},
{
"epoch": 0.7505584512285927,
"grad_norm": 0.16300025582313538,
"learning_rate": 1.5498693796462994e-06,
"loss": 0.5907,
"step": 2520
},
{
"epoch": 0.7565152643335815,
"grad_norm": 0.1716276854276657,
"learning_rate": 1.4807018672724066e-06,
"loss": 0.5824,
"step": 2540
},
{
"epoch": 0.7624720774385704,
"grad_norm": 0.16921989619731903,
"learning_rate": 1.4128440397669301e-06,
"loss": 0.5831,
"step": 2560
},
{
"epoch": 0.7684288905435592,
"grad_norm": 0.16908906400203705,
"learning_rate": 1.3463211499974993e-06,
"loss": 0.5857,
"step": 2580
},
{
"epoch": 0.774385703648548,
"grad_norm": 0.17232823371887207,
"learning_rate": 1.2811579540429836e-06,
"loss": 0.59,
"step": 2600
},
{
"epoch": 0.7803425167535368,
"grad_norm": 0.16725070774555206,
"learning_rate": 1.2173787019806688e-06,
"loss": 0.585,
"step": 2620
},
{
"epoch": 0.7862993298585257,
"grad_norm": 0.1646980494260788,
"learning_rate": 1.1550071288617486e-06,
"loss": 0.5866,
"step": 2640
},
{
"epoch": 0.7922561429635145,
"grad_norm": 0.1682273894548416,
"learning_rate": 1.0940664458784733e-06,
"loss": 0.584,
"step": 2660
},
{
"epoch": 0.7982129560685034,
"grad_norm": 0.16984756290912628,
"learning_rate": 1.0345793317262387e-06,
"loss": 0.5825,
"step": 2680
},
{
"epoch": 0.8041697691734921,
"grad_norm": 0.16755692660808563,
"learning_rate": 9.765679241638649e-07,
"loss": 0.5936,
"step": 2700
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.1689227968454361,
"learning_rate": 9.200538117751595e-07,
"loss": 0.5944,
"step": 2720
},
{
"epoch": 0.8160833953834699,
"grad_norm": 0.17434358596801758,
"learning_rate": 8.650580259348595e-07,
"loss": 0.5942,
"step": 2740
},
{
"epoch": 0.8220402084884587,
"grad_norm": 0.1761932522058487,
"learning_rate": 8.116010329819418e-07,
"loss": 0.5903,
"step": 2760
},
{
"epoch": 0.8279970215934475,
"grad_norm": 0.17172160744667053,
"learning_rate": 7.597027266031992e-07,
"loss": 0.5855,
"step": 2780
},
{
"epoch": 0.8339538346984363,
"grad_norm": 0.1609344631433487,
"learning_rate": 7.093824204299299e-07,
"loss": 0.5859,
"step": 2800
},
{
"epoch": 0.8399106478034252,
"grad_norm": 0.16955158114433289,
"learning_rate": 6.60658840850486e-07,
"loss": 0.5997,
"step": 2820
},
{
"epoch": 0.845867460908414,
"grad_norm": 0.1646656095981598,
"learning_rate": 6.135501200413724e-07,
"loss": 0.5847,
"step": 2840
},
{
"epoch": 0.8518242740134029,
"grad_norm": 0.1641419529914856,
"learning_rate": 5.680737892194654e-07,
"loss": 0.5893,
"step": 2860
},
{
"epoch": 0.8577810871183916,
"grad_norm": 0.17007321119308472,
"learning_rate": 5.242467721178774e-07,
"loss": 0.5892,
"step": 2880
},
{
"epoch": 0.8637379002233805,
"grad_norm": 0.1684292107820511,
"learning_rate": 4.82085378687901e-07,
"loss": 0.5821,
"step": 2900
},
{
"epoch": 0.8696947133283693,
"grad_norm": 0.1711101531982422,
"learning_rate": 4.416052990293596e-07,
"loss": 0.5824,
"step": 2920
},
{
"epoch": 0.8756515264333582,
"grad_norm": 0.16678251326084137,
"learning_rate": 4.028215975516314e-07,
"loss": 0.5886,
"step": 2940
},
{
"epoch": 0.8816083395383469,
"grad_norm": 0.17089425027370453,
"learning_rate": 3.657487073675281e-07,
"loss": 0.5974,
"step": 2960
},
{
"epoch": 0.8875651526433358,
"grad_norm": 0.16579705476760864,
"learning_rate": 3.3040042492210103e-07,
"loss": 0.5817,
"step": 2980
},
{
"epoch": 0.8935219657483247,
"grad_norm": 0.17829783260822296,
"learning_rate": 2.9678990485837553e-07,
"loss": 0.5987,
"step": 3000
},
{
"epoch": 0.8994787788533135,
"grad_norm": 0.15789814293384552,
"learning_rate": 2.6492965512193934e-07,
"loss": 0.5919,
"step": 3020
},
{
"epoch": 0.9054355919583023,
"grad_norm": 0.163740336894989,
"learning_rate": 2.348315323061845e-07,
"loss": 0.6026,
"step": 3040
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.16827383637428284,
"learning_rate": 2.0650673723995097e-07,
"loss": 0.5784,
"step": 3060
},
{
"epoch": 0.91734921816828,
"grad_norm": 0.17383261024951935,
"learning_rate": 1.7996581081920483e-07,
"loss": 0.5834,
"step": 3080
},
{
"epoch": 0.9233060312732688,
"grad_norm": 0.17225860059261322,
"learning_rate": 1.5521863008431725e-07,
"loss": 0.5887,
"step": 3100
},
{
"epoch": 0.9292628443782577,
"grad_norm": 0.165319561958313,
"learning_rate": 1.3227440454437745e-07,
"loss": 0.5851,
"step": 3120
},
{
"epoch": 0.9352196574832464,
"grad_norm": 0.16704821586608887,
"learning_rate": 1.1114167274993537e-07,
"loss": 0.6009,
"step": 3140
},
{
"epoch": 0.9411764705882353,
"grad_norm": 0.172063410282135,
"learning_rate": 9.1828299115434e-08,
"loss": 0.5902,
"step": 3160
},
{
"epoch": 0.9471332836932241,
"grad_norm": 0.1612996906042099,
"learning_rate": 7.434147099251165e-08,
"loss": 0.5862,
"step": 3180
},
{
"epoch": 0.953090096798213,
"grad_norm": 0.1649067997932434,
"learning_rate": 5.86876959952809e-08,
"loss": 0.5837,
"step": 3200
},
{
"epoch": 0.9590469099032017,
"grad_norm": 0.16973944008350372,
"learning_rate": 4.4872799578554724e-08,
"loss": 0.587,
"step": 3220
},
{
"epoch": 0.9650037230081906,
"grad_norm": 0.166892409324646,
"learning_rate": 3.290192286994287e-08,
"loss": 0.591,
"step": 3240
},
{
"epoch": 0.9709605361131795,
"grad_norm": 0.16506598889827728,
"learning_rate": 2.2779520756611163e-08,
"loss": 0.5793,
"step": 3260
},
{
"epoch": 0.9769173492181683,
"grad_norm": 0.16446955502033234,
"learning_rate": 1.450936022742544e-08,
"loss": 0.5977,
"step": 3280
},
{
"epoch": 0.9828741623231572,
"grad_norm": 0.17121705412864685,
"learning_rate": 8.094518971088484e-09,
"loss": 0.5908,
"step": 3300
},
{
"epoch": 0.9888309754281459,
"grad_norm": 0.16913697123527527,
"learning_rate": 3.537384230797303e-09,
"loss": 0.5854,
"step": 3320
},
{
"epoch": 0.9947877885331348,
"grad_norm": 0.17292971909046173,
"learning_rate": 8.396519158465621e-10,
"loss": 0.5844,
"step": 3340
},
{
"epoch": 1.0,
"step": 3358,
"total_flos": 7.06306771390772e+18,
"train_loss": 0.6182214067832852,
"train_runtime": 4522.8123,
"train_samples_per_second": 190.035,
"train_steps_per_second": 0.742
}
],
"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
}