OmniCoT-R1 / trainer_state.json
Eustia1's picture
Upload OmniCoT-R1 model
f6330a4 verified
Raw
History Blame Contribute Delete
47.7 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 2700,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.011120378092855158,
"grad_norm": 17.135483940637105,
"learning_rate": 3.3333333333333335e-07,
"loss": 1.9059,
"step": 10
},
{
"epoch": 0.022240756185710316,
"grad_norm": 11.994831289379958,
"learning_rate": 7.037037037037038e-07,
"loss": 1.8159,
"step": 20
},
{
"epoch": 0.03336113427856547,
"grad_norm": 8.5313328711494,
"learning_rate": 1.074074074074074e-06,
"loss": 1.5319,
"step": 30
},
{
"epoch": 0.04448151237142063,
"grad_norm": 6.462747424009116,
"learning_rate": 1.4444444444444445e-06,
"loss": 1.2454,
"step": 40
},
{
"epoch": 0.055601890464275786,
"grad_norm": 4.459642596260716,
"learning_rate": 1.8148148148148151e-06,
"loss": 0.9937,
"step": 50
},
{
"epoch": 0.06672226855713094,
"grad_norm": 4.328525776915536,
"learning_rate": 2.1851851851851853e-06,
"loss": 0.8582,
"step": 60
},
{
"epoch": 0.0778426466499861,
"grad_norm": 4.384559268280673,
"learning_rate": 2.5555555555555557e-06,
"loss": 0.8171,
"step": 70
},
{
"epoch": 0.08896302474284126,
"grad_norm": 4.509636300928446,
"learning_rate": 2.9259259259259257e-06,
"loss": 0.771,
"step": 80
},
{
"epoch": 0.10008340283569642,
"grad_norm": 4.459641219197237,
"learning_rate": 3.296296296296296e-06,
"loss": 0.7154,
"step": 90
},
{
"epoch": 0.11120378092855157,
"grad_norm": 3.769733484995796,
"learning_rate": 3.6666666666666666e-06,
"loss": 0.7022,
"step": 100
},
{
"epoch": 0.12232415902140673,
"grad_norm": 4.674385409638228,
"learning_rate": 4.037037037037037e-06,
"loss": 0.7072,
"step": 110
},
{
"epoch": 0.13344453711426188,
"grad_norm": 3.9182073599540583,
"learning_rate": 4.407407407407408e-06,
"loss": 0.6981,
"step": 120
},
{
"epoch": 0.14456491520711703,
"grad_norm": 4.083865818135409,
"learning_rate": 4.777777777777778e-06,
"loss": 0.7073,
"step": 130
},
{
"epoch": 0.1556852932999722,
"grad_norm": 4.411895739798695,
"learning_rate": 5.148148148148149e-06,
"loss": 0.6806,
"step": 140
},
{
"epoch": 0.16680567139282734,
"grad_norm": 4.126250148568886,
"learning_rate": 5.518518518518518e-06,
"loss": 0.6597,
"step": 150
},
{
"epoch": 0.17792604948568252,
"grad_norm": 3.73910294544406,
"learning_rate": 5.88888888888889e-06,
"loss": 0.6366,
"step": 160
},
{
"epoch": 0.18904642757853768,
"grad_norm": 3.816207737165413,
"learning_rate": 6.259259259259259e-06,
"loss": 0.6623,
"step": 170
},
{
"epoch": 0.20016680567139283,
"grad_norm": 3.89838850539154,
"learning_rate": 6.62962962962963e-06,
"loss": 0.6669,
"step": 180
},
{
"epoch": 0.211287183764248,
"grad_norm": 4.017500360024897,
"learning_rate": 7e-06,
"loss": 0.6763,
"step": 190
},
{
"epoch": 0.22240756185710314,
"grad_norm": 3.8891130473092286,
"learning_rate": 7.370370370370371e-06,
"loss": 0.6501,
"step": 200
},
{
"epoch": 0.2335279399499583,
"grad_norm": 3.91916746786707,
"learning_rate": 7.74074074074074e-06,
"loss": 0.6554,
"step": 210
},
{
"epoch": 0.24464831804281345,
"grad_norm": 3.81255473616947,
"learning_rate": 8.111111111111112e-06,
"loss": 0.6275,
"step": 220
},
{
"epoch": 0.2557686961356686,
"grad_norm": 4.06589884934052,
"learning_rate": 8.481481481481482e-06,
"loss": 0.6412,
"step": 230
},
{
"epoch": 0.26688907422852376,
"grad_norm": 3.6546289969716765,
"learning_rate": 8.851851851851853e-06,
"loss": 0.6471,
"step": 240
},
{
"epoch": 0.2780094523213789,
"grad_norm": 4.522252456018673,
"learning_rate": 9.222222222222224e-06,
"loss": 0.6384,
"step": 250
},
{
"epoch": 0.28912983041423407,
"grad_norm": 3.773483709709219,
"learning_rate": 9.592592592592593e-06,
"loss": 0.6086,
"step": 260
},
{
"epoch": 0.3002502085070892,
"grad_norm": 4.130916415496019,
"learning_rate": 9.962962962962964e-06,
"loss": 0.6591,
"step": 270
},
{
"epoch": 0.3113705865999444,
"grad_norm": 3.354037894582393,
"learning_rate": 9.999661540018812e-06,
"loss": 0.6232,
"step": 280
},
{
"epoch": 0.32249096469279953,
"grad_norm": 4.140470560915805,
"learning_rate": 9.998491613727142e-06,
"loss": 0.6198,
"step": 290
},
{
"epoch": 0.3336113427856547,
"grad_norm": 3.8764445956881253,
"learning_rate": 9.996486238101964e-06,
"loss": 0.6502,
"step": 300
},
{
"epoch": 0.3447317208785099,
"grad_norm": 3.669747137220483,
"learning_rate": 9.99364574832234e-06,
"loss": 0.6183,
"step": 310
},
{
"epoch": 0.35585209897136505,
"grad_norm": 3.409658743950835,
"learning_rate": 9.98997061914855e-06,
"loss": 0.6349,
"step": 320
},
{
"epoch": 0.3669724770642202,
"grad_norm": 3.5482762919336928,
"learning_rate": 9.985461464842748e-06,
"loss": 0.6218,
"step": 330
},
{
"epoch": 0.37809285515707536,
"grad_norm": 3.4522720428846743,
"learning_rate": 9.98011903906629e-06,
"loss": 0.6387,
"step": 340
},
{
"epoch": 0.3892132332499305,
"grad_norm": 4.219649138271389,
"learning_rate": 9.973944234753763e-06,
"loss": 0.6222,
"step": 350
},
{
"epoch": 0.40033361134278567,
"grad_norm": 3.868744947147,
"learning_rate": 9.966938083963747e-06,
"loss": 0.6403,
"step": 360
},
{
"epoch": 0.4114539894356408,
"grad_norm": 3.324203449235186,
"learning_rate": 9.959101757706308e-06,
"loss": 0.6185,
"step": 370
},
{
"epoch": 0.422574367528496,
"grad_norm": 3.937249704311977,
"learning_rate": 9.950436565747286e-06,
"loss": 0.5915,
"step": 380
},
{
"epoch": 0.43369474562135113,
"grad_norm": 2.8698420025789604,
"learning_rate": 9.940943956389369e-06,
"loss": 0.6076,
"step": 390
},
{
"epoch": 0.4448151237142063,
"grad_norm": 3.495094938718513,
"learning_rate": 9.930625516230026e-06,
"loss": 0.6022,
"step": 400
},
{
"epoch": 0.45593550180706144,
"grad_norm": 3.2310704164066855,
"learning_rate": 9.919482969896333e-06,
"loss": 0.6039,
"step": 410
},
{
"epoch": 0.4670558798999166,
"grad_norm": 3.3159471178285242,
"learning_rate": 9.907518179756704e-06,
"loss": 0.5985,
"step": 420
},
{
"epoch": 0.47817625799277175,
"grad_norm": 3.0225190125159864,
"learning_rate": 9.894733145609623e-06,
"loss": 0.5789,
"step": 430
},
{
"epoch": 0.4892966360856269,
"grad_norm": 4.152698067475681,
"learning_rate": 9.881130004349394e-06,
"loss": 0.6163,
"step": 440
},
{
"epoch": 0.5004170141784821,
"grad_norm": 3.0843397122106486,
"learning_rate": 9.866711029608977e-06,
"loss": 0.6163,
"step": 450
},
{
"epoch": 0.5115373922713372,
"grad_norm": 2.9982138249490817,
"learning_rate": 9.851478631379982e-06,
"loss": 0.6006,
"step": 460
},
{
"epoch": 0.5226577703641924,
"grad_norm": 3.156602829226724,
"learning_rate": 9.835435355609854e-06,
"loss": 0.5958,
"step": 470
},
{
"epoch": 0.5337781484570475,
"grad_norm": 2.993800732078412,
"learning_rate": 9.818583883776343e-06,
"loss": 0.5938,
"step": 480
},
{
"epoch": 0.5448985265499027,
"grad_norm": 3.2963286837870496,
"learning_rate": 9.800927032439322e-06,
"loss": 0.6093,
"step": 490
},
{
"epoch": 0.5560189046427578,
"grad_norm": 3.302092148084543,
"learning_rate": 9.78246775277003e-06,
"loss": 0.6116,
"step": 500
},
{
"epoch": 0.567139282735613,
"grad_norm": 3.2786794147063967,
"learning_rate": 9.763209130057807e-06,
"loss": 0.5851,
"step": 510
},
{
"epoch": 0.5782596608284681,
"grad_norm": 3.145699196243855,
"learning_rate": 9.743154383194422e-06,
"loss": 0.5736,
"step": 520
},
{
"epoch": 0.5893800389213233,
"grad_norm": 2.8385249200393683,
"learning_rate": 9.722306864136064e-06,
"loss": 0.6007,
"step": 530
},
{
"epoch": 0.6005004170141784,
"grad_norm": 2.82953003673451,
"learning_rate": 9.700670057343093e-06,
"loss": 0.5668,
"step": 540
},
{
"epoch": 0.6116207951070336,
"grad_norm": 2.732291498335005,
"learning_rate": 9.678247579197658e-06,
"loss": 0.581,
"step": 550
},
{
"epoch": 0.6227411731998888,
"grad_norm": 2.8470821996613167,
"learning_rate": 9.655043177399236e-06,
"loss": 0.5692,
"step": 560
},
{
"epoch": 0.6338615512927439,
"grad_norm": 2.728109835403873,
"learning_rate": 9.63106073033825e-06,
"loss": 0.5778,
"step": 570
},
{
"epoch": 0.6449819293855991,
"grad_norm": 2.9903971528157833,
"learning_rate": 9.60630424644784e-06,
"loss": 0.5822,
"step": 580
},
{
"epoch": 0.6561023074784542,
"grad_norm": 2.681336224424039,
"learning_rate": 9.58077786353388e-06,
"loss": 0.5885,
"step": 590
},
{
"epoch": 0.6672226855713094,
"grad_norm": 2.940937240658269,
"learning_rate": 9.554485848083389e-06,
"loss": 0.5611,
"step": 600
},
{
"epoch": 0.6783430636641646,
"grad_norm": 2.6240312311662217,
"learning_rate": 9.52743259455143e-06,
"loss": 0.5586,
"step": 610
},
{
"epoch": 0.6894634417570198,
"grad_norm": 2.929869424371381,
"learning_rate": 9.499622624626624e-06,
"loss": 0.5877,
"step": 620
},
{
"epoch": 0.700583819849875,
"grad_norm": 3.2260949562913876,
"learning_rate": 9.471060586475383e-06,
"loss": 0.5768,
"step": 630
},
{
"epoch": 0.7117041979427301,
"grad_norm": 2.4010710022713107,
"learning_rate": 9.441751253965022e-06,
"loss": 0.5612,
"step": 640
},
{
"epoch": 0.7228245760355853,
"grad_norm": 2.7513659573932983,
"learning_rate": 9.411699525865849e-06,
"loss": 0.5774,
"step": 650
},
{
"epoch": 0.7339449541284404,
"grad_norm": 2.721037644213535,
"learning_rate": 9.38091042503239e-06,
"loss": 0.5908,
"step": 660
},
{
"epoch": 0.7450653322212956,
"grad_norm": 2.7139376133624844,
"learning_rate": 9.349389097563858e-06,
"loss": 0.5975,
"step": 670
},
{
"epoch": 0.7561857103141507,
"grad_norm": 2.7069146461484808,
"learning_rate": 9.31714081194404e-06,
"loss": 0.583,
"step": 680
},
{
"epoch": 0.7673060884070059,
"grad_norm": 2.7956232888213597,
"learning_rate": 9.28417095816072e-06,
"loss": 0.5724,
"step": 690
},
{
"epoch": 0.778426466499861,
"grad_norm": 2.7039227850499343,
"learning_rate": 9.25048504680479e-06,
"loss": 0.5714,
"step": 700
},
{
"epoch": 0.7895468445927162,
"grad_norm": 2.8908388236284255,
"learning_rate": 9.216088708149216e-06,
"loss": 0.567,
"step": 710
},
{
"epoch": 0.8006672226855713,
"grad_norm": 2.880449781937227,
"learning_rate": 9.18098769120798e-06,
"loss": 0.5719,
"step": 720
},
{
"epoch": 0.8117876007784265,
"grad_norm": 2.7636319668015554,
"learning_rate": 9.145187862775208e-06,
"loss": 0.5775,
"step": 730
},
{
"epoch": 0.8229079788712816,
"grad_norm": 2.9622544632818397,
"learning_rate": 9.108695206444575e-06,
"loss": 0.5718,
"step": 740
},
{
"epoch": 0.8340283569641368,
"grad_norm": 2.7125967315423054,
"learning_rate": 9.071515821609206e-06,
"loss": 0.5743,
"step": 750
},
{
"epoch": 0.845148735056992,
"grad_norm": 2.716396073201143,
"learning_rate": 9.033655922442235e-06,
"loss": 0.5657,
"step": 760
},
{
"epoch": 0.8562691131498471,
"grad_norm": 2.3353908512216015,
"learning_rate": 8.995121836858156e-06,
"loss": 0.5604,
"step": 770
},
{
"epoch": 0.8673894912427023,
"grad_norm": 2.6241860131781434,
"learning_rate": 8.955920005455175e-06,
"loss": 0.5575,
"step": 780
},
{
"epoch": 0.8785098693355574,
"grad_norm": 2.4926395494435867,
"learning_rate": 8.916056980438723e-06,
"loss": 0.5515,
"step": 790
},
{
"epoch": 0.8896302474284126,
"grad_norm": 2.790041925937432,
"learning_rate": 8.87553942452633e-06,
"loss": 0.5666,
"step": 800
},
{
"epoch": 0.9007506255212677,
"grad_norm": 2.4928250653535717,
"learning_rate": 8.834374109834004e-06,
"loss": 0.5771,
"step": 810
},
{
"epoch": 0.9118710036141229,
"grad_norm": 2.478309201102353,
"learning_rate": 8.792567916744346e-06,
"loss": 0.5467,
"step": 820
},
{
"epoch": 0.922991381706978,
"grad_norm": 2.6170279309225286,
"learning_rate": 8.750127832756556e-06,
"loss": 0.5621,
"step": 830
},
{
"epoch": 0.9341117597998332,
"grad_norm": 2.741182305842299,
"learning_rate": 8.707060951318542e-06,
"loss": 0.5515,
"step": 840
},
{
"epoch": 0.9452321378926883,
"grad_norm": 2.5701026284746473,
"learning_rate": 8.663374470641319e-06,
"loss": 0.5614,
"step": 850
},
{
"epoch": 0.9563525159855435,
"grad_norm": 2.6211495241855527,
"learning_rate": 8.6190756924959e-06,
"loss": 0.5544,
"step": 860
},
{
"epoch": 0.9674728940783986,
"grad_norm": 2.177265028323134,
"learning_rate": 8.574172020992868e-06,
"loss": 0.5672,
"step": 870
},
{
"epoch": 0.9785932721712538,
"grad_norm": 2.4619892606562326,
"learning_rate": 8.528670961344866e-06,
"loss": 0.5475,
"step": 880
},
{
"epoch": 0.989713650264109,
"grad_norm": 2.535258890888737,
"learning_rate": 8.482580118612156e-06,
"loss": 0.5424,
"step": 890
},
{
"epoch": 1.0,
"grad_norm": 4.925821607348586,
"learning_rate": 8.435907196431519e-06,
"loss": 0.5432,
"step": 900
},
{
"epoch": 1.0111203780928553,
"grad_norm": 2.2266984745938805,
"learning_rate": 8.388659995728662e-06,
"loss": 0.4789,
"step": 910
},
{
"epoch": 1.0222407561857103,
"grad_norm": 2.427979206079829,
"learning_rate": 8.340846413414371e-06,
"loss": 0.4477,
"step": 920
},
{
"epoch": 1.0333611342785656,
"grad_norm": 2.3302363292257553,
"learning_rate": 8.292474441064619e-06,
"loss": 0.4688,
"step": 930
},
{
"epoch": 1.0444815123714206,
"grad_norm": 2.412636631834794,
"learning_rate": 8.243552163584851e-06,
"loss": 0.4616,
"step": 940
},
{
"epoch": 1.0556018904642759,
"grad_norm": 2.5397057274192454,
"learning_rate": 8.194087757858668e-06,
"loss": 0.4462,
"step": 950
},
{
"epoch": 1.066722268557131,
"grad_norm": 2.533008589042743,
"learning_rate": 8.144089491381143e-06,
"loss": 0.4514,
"step": 960
},
{
"epoch": 1.0778426466499862,
"grad_norm": 2.488789327094875,
"learning_rate": 8.093565720876994e-06,
"loss": 0.4612,
"step": 970
},
{
"epoch": 1.0889630247428412,
"grad_norm": 2.3603536670160663,
"learning_rate": 8.042524890903828e-06,
"loss": 0.4523,
"step": 980
},
{
"epoch": 1.1000834028356965,
"grad_norm": 2.4044793177414507,
"learning_rate": 7.990975532440728e-06,
"loss": 0.4676,
"step": 990
},
{
"epoch": 1.1112037809285515,
"grad_norm": 2.348733422525813,
"learning_rate": 7.938926261462366e-06,
"loss": 0.4587,
"step": 1000
},
{
"epoch": 1.1223241590214068,
"grad_norm": 2.43658249004558,
"learning_rate": 7.886385777498946e-06,
"loss": 0.4575,
"step": 1010
},
{
"epoch": 1.1334445371142619,
"grad_norm": 2.348130510433153,
"learning_rate": 7.833362862182148e-06,
"loss": 0.4628,
"step": 1020
},
{
"epoch": 1.1445649152071171,
"grad_norm": 2.5293752833344487,
"learning_rate": 7.779866377777367e-06,
"loss": 0.4469,
"step": 1030
},
{
"epoch": 1.1556852932999722,
"grad_norm": 2.2629972396565528,
"learning_rate": 7.725905265702476e-06,
"loss": 0.4621,
"step": 1040
},
{
"epoch": 1.1668056713928274,
"grad_norm": 2.513172182322125,
"learning_rate": 7.671488545033348e-06,
"loss": 0.4615,
"step": 1050
},
{
"epoch": 1.1779260494856825,
"grad_norm": 2.571017185952138,
"learning_rate": 7.616625310996405e-06,
"loss": 0.452,
"step": 1060
},
{
"epoch": 1.1890464275785377,
"grad_norm": 2.407400415942569,
"learning_rate": 7.5613247334484495e-06,
"loss": 0.4384,
"step": 1070
},
{
"epoch": 1.2001668056713928,
"grad_norm": 2.464947469310559,
"learning_rate": 7.505596055344004e-06,
"loss": 0.4638,
"step": 1080
},
{
"epoch": 1.211287183764248,
"grad_norm": 2.4359076610385495,
"learning_rate": 7.449448591190436e-06,
"loss": 0.4779,
"step": 1090
},
{
"epoch": 1.222407561857103,
"grad_norm": 2.4986135840340427,
"learning_rate": 7.392891725491138e-06,
"loss": 0.4654,
"step": 1100
},
{
"epoch": 1.2335279399499584,
"grad_norm": 2.6750568432129413,
"learning_rate": 7.3359349111769965e-06,
"loss": 0.4564,
"step": 1110
},
{
"epoch": 1.2446483180428134,
"grad_norm": 2.3421214676202777,
"learning_rate": 7.278587668026422e-06,
"loss": 0.4754,
"step": 1120
},
{
"epoch": 1.2557686961356687,
"grad_norm": 2.457220092160188,
"learning_rate": 7.220859581074217e-06,
"loss": 0.4501,
"step": 1130
},
{
"epoch": 1.2668890742285237,
"grad_norm": 2.378627397434006,
"learning_rate": 7.162760299009524e-06,
"loss": 0.4491,
"step": 1140
},
{
"epoch": 1.278009452321379,
"grad_norm": 2.240335096743411,
"learning_rate": 7.104299532563146e-06,
"loss": 0.4551,
"step": 1150
},
{
"epoch": 1.289129830414234,
"grad_norm": 2.28852611087441,
"learning_rate": 7.0454870528844885e-06,
"loss": 0.439,
"step": 1160
},
{
"epoch": 1.3002502085070893,
"grad_norm": 2.4649317154302883,
"learning_rate": 6.9863326899084105e-06,
"loss": 0.4679,
"step": 1170
},
{
"epoch": 1.3113705865999443,
"grad_norm": 2.677421841889832,
"learning_rate": 6.9268463307122425e-06,
"loss": 0.4755,
"step": 1180
},
{
"epoch": 1.3224909646927996,
"grad_norm": 2.682694118476393,
"learning_rate": 6.867037917863258e-06,
"loss": 0.4485,
"step": 1190
},
{
"epoch": 1.3336113427856546,
"grad_norm": 2.3780842279263266,
"learning_rate": 6.806917447756874e-06,
"loss": 0.4635,
"step": 1200
},
{
"epoch": 1.34473172087851,
"grad_norm": 2.6323261105171425,
"learning_rate": 6.746494968945847e-06,
"loss": 0.4693,
"step": 1210
},
{
"epoch": 1.355852098971365,
"grad_norm": 2.302266314651215,
"learning_rate": 6.685780580460754e-06,
"loss": 0.4599,
"step": 1220
},
{
"epoch": 1.3669724770642202,
"grad_norm": 2.5315952060668976,
"learning_rate": 6.624784430122041e-06,
"loss": 0.4518,
"step": 1230
},
{
"epoch": 1.3780928551570755,
"grad_norm": 2.6470800808239425,
"learning_rate": 6.56351671284391e-06,
"loss": 0.4626,
"step": 1240
},
{
"epoch": 1.3892132332499305,
"grad_norm": 2.5471713455157508,
"learning_rate": 6.501987668930343e-06,
"loss": 0.4563,
"step": 1250
},
{
"epoch": 1.4003336113427856,
"grad_norm": 2.14907267752596,
"learning_rate": 6.440207582363529e-06,
"loss": 0.447,
"step": 1260
},
{
"epoch": 1.4114539894356408,
"grad_norm": 2.478643654460266,
"learning_rate": 6.378186779084996e-06,
"loss": 0.4658,
"step": 1270
},
{
"epoch": 1.422574367528496,
"grad_norm": 2.797985035291058,
"learning_rate": 6.315935625269739e-06,
"loss": 0.4719,
"step": 1280
},
{
"epoch": 1.4336947456213511,
"grad_norm": 2.6006382925345877,
"learning_rate": 6.253464525593605e-06,
"loss": 0.4497,
"step": 1290
},
{
"epoch": 1.4448151237142062,
"grad_norm": 2.3759599361813946,
"learning_rate": 6.190783921494255e-06,
"loss": 0.4453,
"step": 1300
},
{
"epoch": 1.4559355018070614,
"grad_norm": 2.3513510228163392,
"learning_rate": 6.127904289425979e-06,
"loss": 0.4323,
"step": 1310
},
{
"epoch": 1.4670558798999167,
"grad_norm": 2.527290429496959,
"learning_rate": 6.064836139108664e-06,
"loss": 0.4564,
"step": 1320
},
{
"epoch": 1.4781762579927717,
"grad_norm": 2.409684354160994,
"learning_rate": 6.001590011771188e-06,
"loss": 0.4441,
"step": 1330
},
{
"epoch": 1.4892966360856268,
"grad_norm": 2.2887885185453847,
"learning_rate": 5.938176478389557e-06,
"loss": 0.4608,
"step": 1340
},
{
"epoch": 1.500417014178482,
"grad_norm": 2.280496810998306,
"learning_rate": 5.8746061379200796e-06,
"loss": 0.4688,
"step": 1350
},
{
"epoch": 1.5115373922713373,
"grad_norm": 2.7461463582565138,
"learning_rate": 5.810889615527839e-06,
"loss": 0.455,
"step": 1360
},
{
"epoch": 1.5226577703641924,
"grad_norm": 2.7174833473654547,
"learning_rate": 5.747037560810809e-06,
"loss": 0.4454,
"step": 1370
},
{
"epoch": 1.5337781484570474,
"grad_norm": 2.3594202410324385,
"learning_rate": 5.683060646019878e-06,
"loss": 0.4526,
"step": 1380
},
{
"epoch": 1.5448985265499027,
"grad_norm": 2.487208425433172,
"learning_rate": 5.618969564275083e-06,
"loss": 0.455,
"step": 1390
},
{
"epoch": 1.556018904642758,
"grad_norm": 2.4665224035083497,
"learning_rate": 5.554775027778352e-06,
"loss": 0.4568,
"step": 1400
},
{
"epoch": 1.567139282735613,
"grad_norm": 2.4959214278689212,
"learning_rate": 5.49048776602308e-06,
"loss": 0.4625,
"step": 1410
},
{
"epoch": 1.578259660828468,
"grad_norm": 2.6454517892459752,
"learning_rate": 5.426118524000784e-06,
"loss": 0.4581,
"step": 1420
},
{
"epoch": 1.5893800389213233,
"grad_norm": 2.489711134640338,
"learning_rate": 5.361678060405184e-06,
"loss": 0.4472,
"step": 1430
},
{
"epoch": 1.6005004170141786,
"grad_norm": 2.6206710174604373,
"learning_rate": 5.297177145833997e-06,
"loss": 0.4426,
"step": 1440
},
{
"epoch": 1.6116207951070336,
"grad_norm": 2.2427644912982965,
"learning_rate": 5.232626560988735e-06,
"loss": 0.4493,
"step": 1450
},
{
"epoch": 1.6227411731998886,
"grad_norm": 2.5822066750100867,
"learning_rate": 5.168037094872809e-06,
"loss": 0.4552,
"step": 1460
},
{
"epoch": 1.633861551292744,
"grad_norm": 2.361047074606954,
"learning_rate": 5.103419542988264e-06,
"loss": 0.4297,
"step": 1470
},
{
"epoch": 1.6449819293855992,
"grad_norm": 2.2851228310327225,
"learning_rate": 5.038784705531402e-06,
"loss": 0.4417,
"step": 1480
},
{
"epoch": 1.6561023074784542,
"grad_norm": 2.3068196852688665,
"learning_rate": 4.974143385587644e-06,
"loss": 0.4281,
"step": 1490
},
{
"epoch": 1.6672226855713093,
"grad_norm": 2.5952180527180424,
"learning_rate": 4.909506387325894e-06,
"loss": 0.4378,
"step": 1500
},
{
"epoch": 1.6783430636641645,
"grad_norm": 2.485419768058454,
"learning_rate": 4.844884514192732e-06,
"loss": 0.4465,
"step": 1510
},
{
"epoch": 1.6894634417570198,
"grad_norm": 2.4545613254593537,
"learning_rate": 4.780288567106715e-06,
"loss": 0.4315,
"step": 1520
},
{
"epoch": 1.700583819849875,
"grad_norm": 2.3848358914543972,
"learning_rate": 4.715729342653114e-06,
"loss": 0.4466,
"step": 1530
},
{
"epoch": 1.71170419794273,
"grad_norm": 2.2852753758236166,
"learning_rate": 4.651217631279374e-06,
"loss": 0.4476,
"step": 1540
},
{
"epoch": 1.7228245760355851,
"grad_norm": 2.2956402356015326,
"learning_rate": 4.586764215491583e-06,
"loss": 0.4631,
"step": 1550
},
{
"epoch": 1.7339449541284404,
"grad_norm": 2.3602668238453863,
"learning_rate": 4.522379868052292e-06,
"loss": 0.4283,
"step": 1560
},
{
"epoch": 1.7450653322212957,
"grad_norm": 2.6227232003978282,
"learning_rate": 4.458075350179948e-06,
"loss": 0.4385,
"step": 1570
},
{
"epoch": 1.7561857103141507,
"grad_norm": 2.406344468203441,
"learning_rate": 4.3938614097502565e-06,
"loss": 0.4253,
"step": 1580
},
{
"epoch": 1.7673060884070058,
"grad_norm": 2.587804438277781,
"learning_rate": 4.3297487794997895e-06,
"loss": 0.4434,
"step": 1590
},
{
"epoch": 1.778426466499861,
"grad_norm": 2.28205972431158,
"learning_rate": 4.265748175232105e-06,
"loss": 0.4573,
"step": 1600
},
{
"epoch": 1.7895468445927163,
"grad_norm": 2.48717276910565,
"learning_rate": 4.201870294026704e-06,
"loss": 0.4074,
"step": 1610
},
{
"epoch": 1.8006672226855713,
"grad_norm": 2.528211660489936,
"learning_rate": 4.138125812451117e-06,
"loss": 0.4341,
"step": 1620
},
{
"epoch": 1.8117876007784264,
"grad_norm": 2.1731304371314986,
"learning_rate": 4.074525384776428e-06,
"loss": 0.4409,
"step": 1630
},
{
"epoch": 1.8229079788712816,
"grad_norm": 2.186242836383946,
"learning_rate": 4.011079641196505e-06,
"loss": 0.4151,
"step": 1640
},
{
"epoch": 1.834028356964137,
"grad_norm": 2.3557356336739854,
"learning_rate": 3.947799186051273e-06,
"loss": 0.448,
"step": 1650
},
{
"epoch": 1.845148735056992,
"grad_norm": 2.2983289228420407,
"learning_rate": 3.884694596054304e-06,
"loss": 0.4363,
"step": 1660
},
{
"epoch": 1.856269113149847,
"grad_norm": 2.520946028881087,
"learning_rate": 3.821776418525013e-06,
"loss": 0.4315,
"step": 1670
},
{
"epoch": 1.8673894912427023,
"grad_norm": 2.405218987495231,
"learning_rate": 3.759055169625784e-06,
"loss": 0.4239,
"step": 1680
},
{
"epoch": 1.8785098693355575,
"grad_norm": 2.3323182810580656,
"learning_rate": 3.6965413326042933e-06,
"loss": 0.4152,
"step": 1690
},
{
"epoch": 1.8896302474284126,
"grad_norm": 2.404917026117654,
"learning_rate": 3.6342453560413336e-06,
"loss": 0.4241,
"step": 1700
},
{
"epoch": 1.9007506255212676,
"grad_norm": 2.5779583464543263,
"learning_rate": 3.5721776521044384e-06,
"loss": 0.4431,
"step": 1710
},
{
"epoch": 1.9118710036141229,
"grad_norm": 2.3564984929454167,
"learning_rate": 3.51034859480759e-06,
"loss": 0.4279,
"step": 1720
},
{
"epoch": 1.9229913817069781,
"grad_norm": 2.7693011575140503,
"learning_rate": 3.4487685182773028e-06,
"loss": 0.4139,
"step": 1730
},
{
"epoch": 1.9341117597998332,
"grad_norm": 2.380814176259524,
"learning_rate": 3.3874477150253703e-06,
"loss": 0.4405,
"step": 1740
},
{
"epoch": 1.9452321378926882,
"grad_norm": 2.32367799498033,
"learning_rate": 3.3263964342285795e-06,
"loss": 0.4459,
"step": 1750
},
{
"epoch": 1.9563525159855435,
"grad_norm": 2.7514454655257885,
"learning_rate": 3.265624880015652e-06,
"loss": 0.4299,
"step": 1760
},
{
"epoch": 1.9674728940783988,
"grad_norm": 2.5878093856496154,
"learning_rate": 3.2051432097617286e-06,
"loss": 0.4282,
"step": 1770
},
{
"epoch": 1.9785932721712538,
"grad_norm": 2.432888585317069,
"learning_rate": 3.1449615323906657e-06,
"loss": 0.4308,
"step": 1780
},
{
"epoch": 1.9897136502641088,
"grad_norm": 2.4956968658426453,
"learning_rate": 3.085089906685419e-06,
"loss": 0.4331,
"step": 1790
},
{
"epoch": 2.0,
"grad_norm": 4.301781719897612,
"learning_rate": 3.025538339606816e-06,
"loss": 0.4072,
"step": 1800
},
{
"epoch": 2.0111203780928553,
"grad_norm": 2.3213702571346078,
"learning_rate": 2.966316784621e-06,
"loss": 0.2976,
"step": 1810
},
{
"epoch": 2.0222407561857105,
"grad_norm": 2.4721841096175274,
"learning_rate": 2.9074351400357904e-06,
"loss": 0.2861,
"step": 1820
},
{
"epoch": 2.0333611342785654,
"grad_norm": 2.4045514282764513,
"learning_rate": 2.8489032473462874e-06,
"loss": 0.2766,
"step": 1830
},
{
"epoch": 2.0444815123714206,
"grad_norm": 2.1703378206858828,
"learning_rate": 2.790730889589962e-06,
"loss": 0.2849,
"step": 1840
},
{
"epoch": 2.055601890464276,
"grad_norm": 2.4672160825071923,
"learning_rate": 2.732927789711506e-06,
"loss": 0.2793,
"step": 1850
},
{
"epoch": 2.066722268557131,
"grad_norm": 2.3333136431851864,
"learning_rate": 2.6755036089377464e-06,
"loss": 0.2744,
"step": 1860
},
{
"epoch": 2.077842646649986,
"grad_norm": 2.5089718915008326,
"learning_rate": 2.6184679451628587e-06,
"loss": 0.2747,
"step": 1870
},
{
"epoch": 2.0889630247428412,
"grad_norm": 2.334889544087653,
"learning_rate": 2.561830331344176e-06,
"loss": 0.2762,
"step": 1880
},
{
"epoch": 2.1000834028356965,
"grad_norm": 2.289500383346266,
"learning_rate": 2.50560023390885e-06,
"loss": 0.2582,
"step": 1890
},
{
"epoch": 2.1112037809285518,
"grad_norm": 2.6192871425721993,
"learning_rate": 2.4497870511716237e-06,
"loss": 0.2732,
"step": 1900
},
{
"epoch": 2.1223241590214066,
"grad_norm": 2.366840733586268,
"learning_rate": 2.3944001117639957e-06,
"loss": 0.2706,
"step": 1910
},
{
"epoch": 2.133444537114262,
"grad_norm": 2.437470761530495,
"learning_rate": 2.3394486730750334e-06,
"loss": 0.2809,
"step": 1920
},
{
"epoch": 2.144564915207117,
"grad_norm": 2.6262383785480488,
"learning_rate": 2.284941919704085e-06,
"loss": 0.2782,
"step": 1930
},
{
"epoch": 2.1556852932999724,
"grad_norm": 2.5525410585441493,
"learning_rate": 2.2308889619256687e-06,
"loss": 0.2782,
"step": 1940
},
{
"epoch": 2.166805671392827,
"grad_norm": 2.601311914229026,
"learning_rate": 2.1772988341667754e-06,
"loss": 0.2823,
"step": 1950
},
{
"epoch": 2.1779260494856825,
"grad_norm": 2.3592055040841933,
"learning_rate": 2.1241804934968558e-06,
"loss": 0.2693,
"step": 1960
},
{
"epoch": 2.1890464275785377,
"grad_norm": 2.4145342750346757,
"learning_rate": 2.071542818130726e-06,
"loss": 0.2843,
"step": 1970
},
{
"epoch": 2.200166805671393,
"grad_norm": 2.4403194127700605,
"learning_rate": 2.0193946059446644e-06,
"loss": 0.2756,
"step": 1980
},
{
"epoch": 2.211287183764248,
"grad_norm": 2.2682353911922593,
"learning_rate": 1.9677445730059348e-06,
"loss": 0.2721,
"step": 1990
},
{
"epoch": 2.222407561857103,
"grad_norm": 2.2803039934931255,
"learning_rate": 1.916601352115972e-06,
"loss": 0.2683,
"step": 2000
},
{
"epoch": 2.2335279399499584,
"grad_norm": 2.2121843753304904,
"learning_rate": 1.8659734913675059e-06,
"loss": 0.2753,
"step": 2010
},
{
"epoch": 2.2446483180428136,
"grad_norm": 2.2916339346740884,
"learning_rate": 1.8158694527158205e-06,
"loss": 0.2686,
"step": 2020
},
{
"epoch": 2.2557686961356684,
"grad_norm": 2.3487144709512173,
"learning_rate": 1.7662976105644214e-06,
"loss": 0.266,
"step": 2030
},
{
"epoch": 2.2668890742285237,
"grad_norm": 2.710385504354685,
"learning_rate": 1.7172662503653376e-06,
"loss": 0.2787,
"step": 2040
},
{
"epoch": 2.278009452321379,
"grad_norm": 2.7334271906718044,
"learning_rate": 1.6687835672342895e-06,
"loss": 0.2822,
"step": 2050
},
{
"epoch": 2.2891298304142342,
"grad_norm": 2.628705607579475,
"learning_rate": 1.6208576645809515e-06,
"loss": 0.2692,
"step": 2060
},
{
"epoch": 2.300250208507089,
"grad_norm": 2.336117586269706,
"learning_rate": 1.573496552754552e-06,
"loss": 0.2673,
"step": 2070
},
{
"epoch": 2.3113705865999443,
"grad_norm": 2.5128867892733036,
"learning_rate": 1.5267081477050132e-06,
"loss": 0.2623,
"step": 2080
},
{
"epoch": 2.3224909646927996,
"grad_norm": 2.3669464578963524,
"learning_rate": 1.4805002696598853e-06,
"loss": 0.2692,
"step": 2090
},
{
"epoch": 2.333611342785655,
"grad_norm": 2.611177739145231,
"learning_rate": 1.434880641817263e-06,
"loss": 0.2744,
"step": 2100
},
{
"epoch": 2.34473172087851,
"grad_norm": 2.542366353544081,
"learning_rate": 1.3898568890549335e-06,
"loss": 0.2904,
"step": 2110
},
{
"epoch": 2.355852098971365,
"grad_norm": 2.3027230699242134,
"learning_rate": 1.3454365366559513e-06,
"loss": 0.2733,
"step": 2120
},
{
"epoch": 2.36697247706422,
"grad_norm": 2.4288855630538566,
"learning_rate": 1.3016270090508575e-06,
"loss": 0.271,
"step": 2130
},
{
"epoch": 2.3780928551570755,
"grad_norm": 2.3640075562246796,
"learning_rate": 1.2584356285767652e-06,
"loss": 0.2668,
"step": 2140
},
{
"epoch": 2.3892132332499303,
"grad_norm": 2.5472545226747645,
"learning_rate": 1.2158696142534943e-06,
"loss": 0.2711,
"step": 2150
},
{
"epoch": 2.4003336113427856,
"grad_norm": 2.409981303645627,
"learning_rate": 1.1739360805769856e-06,
"loss": 0.2709,
"step": 2160
},
{
"epoch": 2.411453989435641,
"grad_norm": 2.7186825736208773,
"learning_rate": 1.132642036330181e-06,
"loss": 0.2654,
"step": 2170
},
{
"epoch": 2.422574367528496,
"grad_norm": 2.5628089937562897,
"learning_rate": 1.0919943834115743e-06,
"loss": 0.2638,
"step": 2180
},
{
"epoch": 2.4336947456213514,
"grad_norm": 2.543969475835257,
"learning_rate": 1.051999915681629e-06,
"loss": 0.2668,
"step": 2190
},
{
"epoch": 2.444815123714206,
"grad_norm": 2.5227852893318885,
"learning_rate": 1.0126653178272422e-06,
"loss": 0.2579,
"step": 2200
},
{
"epoch": 2.4559355018070614,
"grad_norm": 2.5938636434776536,
"learning_rate": 9.739971642444745e-07,
"loss": 0.2766,
"step": 2210
},
{
"epoch": 2.4670558798999167,
"grad_norm": 2.547142137212308,
"learning_rate": 9.360019179396945e-07,
"loss": 0.2767,
"step": 2220
},
{
"epoch": 2.4781762579927715,
"grad_norm": 2.494385047103373,
"learning_rate": 8.98685929449355e-07,
"loss": 0.2677,
"step": 2230
},
{
"epoch": 2.489296636085627,
"grad_norm": 2.4890527753607707,
"learning_rate": 8.620554357785599e-07,
"loss": 0.2648,
"step": 2240
},
{
"epoch": 2.500417014178482,
"grad_norm": 2.4808720880679562,
"learning_rate": 8.261165593586145e-07,
"loss": 0.2758,
"step": 2250
},
{
"epoch": 2.5115373922713373,
"grad_norm": 2.3620030763685884,
"learning_rate": 7.908753070237124e-07,
"loss": 0.259,
"step": 2260
},
{
"epoch": 2.5226577703641926,
"grad_norm": 2.2006624664507206,
"learning_rate": 7.563375690069602e-07,
"loss": 0.2727,
"step": 2270
},
{
"epoch": 2.5337781484570474,
"grad_norm": 2.6239231677849846,
"learning_rate": 7.225091179558747e-07,
"loss": 0.2691,
"step": 2280
},
{
"epoch": 2.5448985265499027,
"grad_norm": 2.48553690226059,
"learning_rate": 6.893956079675452e-07,
"loss": 0.2709,
"step": 2290
},
{
"epoch": 2.556018904642758,
"grad_norm": 2.3773094991790544,
"learning_rate": 6.570025736436031e-07,
"loss": 0.2574,
"step": 2300
},
{
"epoch": 2.5671392827356128,
"grad_norm": 2.4584450821756056,
"learning_rate": 6.253354291651748e-07,
"loss": 0.2592,
"step": 2310
},
{
"epoch": 2.578259660828468,
"grad_norm": 2.3737590082392437,
"learning_rate": 5.943994673879405e-07,
"loss": 0.2586,
"step": 2320
},
{
"epoch": 2.5893800389213233,
"grad_norm": 2.4740467944491984,
"learning_rate": 5.641998589574993e-07,
"loss": 0.2774,
"step": 2330
},
{
"epoch": 2.6005004170141786,
"grad_norm": 2.5522188488030535,
"learning_rate": 5.347416514451343e-07,
"loss": 0.26,
"step": 2340
},
{
"epoch": 2.611620795107034,
"grad_norm": 2.4727273458027756,
"learning_rate": 5.06029768504166e-07,
"loss": 0.2645,
"step": 2350
},
{
"epoch": 2.6227411731998886,
"grad_norm": 2.478352430608564,
"learning_rate": 4.780690090470091e-07,
"loss": 0.2651,
"step": 2360
},
{
"epoch": 2.633861551292744,
"grad_norm": 2.4128960027740405,
"learning_rate": 4.5086404644308634e-07,
"loss": 0.2609,
"step": 2370
},
{
"epoch": 2.644981929385599,
"grad_norm": 2.285962091201634,
"learning_rate": 4.2441942773771114e-07,
"loss": 0.2565,
"step": 2380
},
{
"epoch": 2.656102307478454,
"grad_norm": 2.625675027646378,
"learning_rate": 3.987395728921006e-07,
"loss": 0.2701,
"step": 2390
},
{
"epoch": 2.6672226855713093,
"grad_norm": 2.570751252477893,
"learning_rate": 3.7382877404461784e-07,
"loss": 0.2628,
"step": 2400
},
{
"epoch": 2.6783430636641645,
"grad_norm": 2.3173778417768776,
"learning_rate": 3.496911947933845e-07,
"loss": 0.2686,
"step": 2410
},
{
"epoch": 2.68946344175702,
"grad_norm": 2.1748188948146185,
"learning_rate": 3.263308695003753e-07,
"loss": 0.2597,
"step": 2420
},
{
"epoch": 2.700583819849875,
"grad_norm": 2.6396680676764395,
"learning_rate": 3.037517026171144e-07,
"loss": 0.2702,
"step": 2430
},
{
"epoch": 2.71170419794273,
"grad_norm": 2.4557556115410053,
"learning_rate": 2.819574680320825e-07,
"loss": 0.2686,
"step": 2440
},
{
"epoch": 2.722824576035585,
"grad_norm": 2.4566452708372695,
"learning_rate": 2.609518084399493e-07,
"loss": 0.2528,
"step": 2450
},
{
"epoch": 2.7339449541284404,
"grad_norm": 2.614139148356986,
"learning_rate": 2.407382347327325e-07,
"loss": 0.2622,
"step": 2460
},
{
"epoch": 2.7450653322212957,
"grad_norm": 2.8330255731260805,
"learning_rate": 2.2132012541298542e-07,
"loss": 0.2683,
"step": 2470
},
{
"epoch": 2.756185710314151,
"grad_norm": 2.485512869711635,
"learning_rate": 2.0270072602911416e-07,
"loss": 0.264,
"step": 2480
},
{
"epoch": 2.7673060884070058,
"grad_norm": 2.3317333072876316,
"learning_rate": 1.8488314863291557e-07,
"loss": 0.2592,
"step": 2490
},
{
"epoch": 2.778426466499861,
"grad_norm": 2.526370069288075,
"learning_rate": 1.6787037125942706e-07,
"loss": 0.2694,
"step": 2500
},
{
"epoch": 2.7895468445927163,
"grad_norm": 2.5654225148437466,
"learning_rate": 1.5166523742917716e-07,
"loss": 0.2607,
"step": 2510
},
{
"epoch": 2.800667222685571,
"grad_norm": 2.689725091294952,
"learning_rate": 1.3627045567291998e-07,
"loss": 0.2789,
"step": 2520
},
{
"epoch": 2.8117876007784264,
"grad_norm": 2.4963578497355647,
"learning_rate": 1.2168859907892904e-07,
"loss": 0.2664,
"step": 2530
},
{
"epoch": 2.8229079788712816,
"grad_norm": 2.2974435458505496,
"learning_rate": 1.0792210486293087e-07,
"loss": 0.273,
"step": 2540
},
{
"epoch": 2.834028356964137,
"grad_norm": 2.568391031502576,
"learning_rate": 9.497327396075062e-08,
"loss": 0.2634,
"step": 2550
},
{
"epoch": 2.845148735056992,
"grad_norm": 2.396915019486478,
"learning_rate": 8.284427064372769e-08,
"loss": 0.2624,
"step": 2560
},
{
"epoch": 2.856269113149847,
"grad_norm": 2.2182586178203354,
"learning_rate": 7.153712215698327e-08,
"loss": 0.266,
"step": 2570
},
{
"epoch": 2.8673894912427023,
"grad_norm": 2.23150882157135,
"learning_rate": 6.105371838058594e-08,
"loss": 0.2596,
"step": 2580
},
{
"epoch": 2.8785098693355575,
"grad_norm": 2.55386888641815,
"learning_rate": 5.139581151367312e-08,
"loss": 0.2729,
"step": 2590
},
{
"epoch": 2.8896302474284123,
"grad_norm": 2.4124201954811104,
"learning_rate": 4.256501578158989e-08,
"loss": 0.2755,
"step": 2600
},
{
"epoch": 2.9007506255212676,
"grad_norm": 2.4180970079660433,
"learning_rate": 3.456280716608973e-08,
"loss": 0.2652,
"step": 2610
},
{
"epoch": 2.911871003614123,
"grad_norm": 2.3157036349557103,
"learning_rate": 2.7390523158633552e-08,
"loss": 0.2533,
"step": 2620
},
{
"epoch": 2.922991381706978,
"grad_norm": 2.500075652578648,
"learning_rate": 2.104936253684353e-08,
"loss": 0.2506,
"step": 2630
},
{
"epoch": 2.9341117597998334,
"grad_norm": 2.3578291845338257,
"learning_rate": 1.554038516414058e-08,
"loss": 0.2697,
"step": 2640
},
{
"epoch": 2.9452321378926882,
"grad_norm": 2.527852943160925,
"learning_rate": 1.0864511812594958e-08,
"loss": 0.2724,
"step": 2650
},
{
"epoch": 2.9563525159855435,
"grad_norm": 2.4088273653738925,
"learning_rate": 7.022524009029919e-09,
"loss": 0.2728,
"step": 2660
},
{
"epoch": 2.9674728940783988,
"grad_norm": 2.308258170950676,
"learning_rate": 4.015063904398986e-09,
"loss": 0.2603,
"step": 2670
},
{
"epoch": 2.9785932721712536,
"grad_norm": 2.346171694747588,
"learning_rate": 1.8426341664529168e-09,
"loss": 0.2694,
"step": 2680
},
{
"epoch": 2.989713650264109,
"grad_norm": 2.3827134282450317,
"learning_rate": 5.055978957257957e-10,
"loss": 0.2607,
"step": 2690
},
{
"epoch": 3.0,
"grad_norm": 3.767157904407931,
"learning_rate": 4.178564846912814e-12,
"loss": 0.2423,
"step": 2700
},
{
"epoch": 3.0,
"step": 2700,
"total_flos": 148554952638464.0,
"train_loss": 0.4593236620337875,
"train_runtime": 13320.8967,
"train_samples_per_second": 3.24,
"train_steps_per_second": 0.203
}
],
"logging_steps": 10,
"max_steps": 2700,
"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": true
},
"attributes": {}
}
},
"total_flos": 148554952638464.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}