Files
AetherCuTF-Ko-2-unique-Seed…/trainer_state.json

1136 lines
30 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": 39202,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006377266321018832,
"grad_norm": 2.5775675773620605,
"learning_rate": 3.175210405508799e-05,
"loss": 3.65084814453125,
"step": 250
},
{
"epoch": 0.012754532642037663,
"grad_norm": 3.2221262454986572,
"learning_rate": 6.363172660035706e-05,
"loss": 1.99697509765625,
"step": 500
},
{
"epoch": 0.019131798963056498,
"grad_norm": 2.5724740028381348,
"learning_rate": 9.551134914562612e-05,
"loss": 1.60589306640625,
"step": 750
},
{
"epoch": 0.025509065284075327,
"grad_norm": 1.6085478067398071,
"learning_rate": 0.00012739097169089516,
"loss": 1.444733154296875,
"step": 1000
},
{
"epoch": 0.03188633160509416,
"grad_norm": 0.7356601357460022,
"learning_rate": 0.00015927059423616425,
"loss": 1.33519287109375,
"step": 1250
},
{
"epoch": 0.038263597926112995,
"grad_norm": 0.558532178401947,
"learning_rate": 0.0001911502167814333,
"loss": 1.224727783203125,
"step": 1500
},
{
"epoch": 0.044640864247131824,
"grad_norm": 0.3695690631866455,
"learning_rate": 0.00022302983932670237,
"loss": 1.12808984375,
"step": 1750
},
{
"epoch": 0.05101813056815065,
"grad_norm": 0.271986722946167,
"learning_rate": 0.00025490946187197144,
"loss": 1.074133544921875,
"step": 2000
},
{
"epoch": 0.05739539688916949,
"grad_norm": 0.2766456604003906,
"learning_rate": 0.00028678908441724047,
"loss": 1.035495849609375,
"step": 2250
},
{
"epoch": 0.06377266321018832,
"grad_norm": 0.21983842551708221,
"learning_rate": 0.00031866870696250956,
"loss": 1.0116007080078124,
"step": 2500
},
{
"epoch": 0.07014992953120715,
"grad_norm": 0.2180081009864807,
"learning_rate": 0.00035054832950777865,
"loss": 0.99118798828125,
"step": 2750
},
{
"epoch": 0.07652719585222599,
"grad_norm": 0.18447472155094147,
"learning_rate": 0.0003824279520530477,
"loss": 0.9717527465820313,
"step": 3000
},
{
"epoch": 0.08290446217324482,
"grad_norm": 0.15918593108654022,
"learning_rate": 0.0004143075745983167,
"loss": 0.96016064453125,
"step": 3250
},
{
"epoch": 0.08928172849426365,
"grad_norm": 0.15337951481342316,
"learning_rate": 0.00044618719714358587,
"loss": 0.94656005859375,
"step": 3500
},
{
"epoch": 0.09565899481528248,
"grad_norm": 0.1397962123155594,
"learning_rate": 0.0004780668196888549,
"loss": 0.933742431640625,
"step": 3750
},
{
"epoch": 0.1020362611363013,
"grad_norm": 0.15751700103282928,
"learning_rate": 0.0004988945891556362,
"loss": 0.9267822265625,
"step": 4000
},
{
"epoch": 0.10841352745732015,
"grad_norm": 0.1273936778306961,
"learning_rate": 0.0004953516056801111,
"loss": 0.9156344604492187,
"step": 4250
},
{
"epoch": 0.11479079377833898,
"grad_norm": 0.12347358465194702,
"learning_rate": 0.0004918086222045861,
"loss": 0.9085472412109376,
"step": 4500
},
{
"epoch": 0.12116806009935781,
"grad_norm": 0.11907054483890533,
"learning_rate": 0.000488265638729061,
"loss": 0.8965045166015625,
"step": 4750
},
{
"epoch": 0.12754532642037664,
"grad_norm": 0.12186691164970398,
"learning_rate": 0.00048472265525353593,
"loss": 0.8894534912109375,
"step": 5000
},
{
"epoch": 0.13392259274139548,
"grad_norm": 0.11302649229764938,
"learning_rate": 0.00048117967177801086,
"loss": 0.8835716552734375,
"step": 5250
},
{
"epoch": 0.1402998590624143,
"grad_norm": 0.11347561329603195,
"learning_rate": 0.00047763668830248573,
"loss": 0.8759080200195313,
"step": 5500
},
{
"epoch": 0.14667712538343314,
"grad_norm": 0.10620150715112686,
"learning_rate": 0.00047409370482696066,
"loss": 0.86854833984375,
"step": 5750
},
{
"epoch": 0.15305439170445198,
"grad_norm": 0.10604240000247955,
"learning_rate": 0.00047055072135143564,
"loss": 0.865083984375,
"step": 6000
},
{
"epoch": 0.1594316580254708,
"grad_norm": 0.10537559539079666,
"learning_rate": 0.00046700773787591057,
"loss": 0.860940185546875,
"step": 6250
},
{
"epoch": 0.16580892434648964,
"grad_norm": 0.10088460892438889,
"learning_rate": 0.0004634647544003855,
"loss": 0.8544019775390626,
"step": 6500
},
{
"epoch": 0.17218619066750845,
"grad_norm": 0.09970396012067795,
"learning_rate": 0.0004599217709248604,
"loss": 0.8502755126953125,
"step": 6750
},
{
"epoch": 0.1785634569885273,
"grad_norm": 0.10199850797653198,
"learning_rate": 0.00045637878744933535,
"loss": 0.845378173828125,
"step": 7000
},
{
"epoch": 0.18494072330954614,
"grad_norm": 0.09216546267271042,
"learning_rate": 0.0004528358039738103,
"loss": 0.8421669311523438,
"step": 7250
},
{
"epoch": 0.19131798963056496,
"grad_norm": 0.10916922241449356,
"learning_rate": 0.0004492928204982852,
"loss": 0.841639892578125,
"step": 7500
},
{
"epoch": 0.1976952559515838,
"grad_norm": 0.09349249303340912,
"learning_rate": 0.00044574983702276013,
"loss": 0.8376875610351563,
"step": 7750
},
{
"epoch": 0.2040725222726026,
"grad_norm": 0.09555172175168991,
"learning_rate": 0.00044220685354723506,
"loss": 0.8310078735351563,
"step": 8000
},
{
"epoch": 0.21044978859362146,
"grad_norm": 0.09085638076066971,
"learning_rate": 0.00043866387007171,
"loss": 0.8307603149414062,
"step": 8250
},
{
"epoch": 0.2168270549146403,
"grad_norm": 0.09479482471942902,
"learning_rate": 0.0004351208865961849,
"loss": 0.8274447631835937,
"step": 8500
},
{
"epoch": 0.22320432123565911,
"grad_norm": 0.0888662189245224,
"learning_rate": 0.00043157790312065985,
"loss": 0.8251286010742187,
"step": 8750
},
{
"epoch": 0.22958158755667796,
"grad_norm": 0.09127531945705414,
"learning_rate": 0.0004280349196451348,
"loss": 0.8239066772460938,
"step": 9000
},
{
"epoch": 0.2359588538776968,
"grad_norm": 0.09323842078447342,
"learning_rate": 0.00042449193616960975,
"loss": 0.821473876953125,
"step": 9250
},
{
"epoch": 0.24233612019871562,
"grad_norm": 0.08932389318943024,
"learning_rate": 0.00042094895269408463,
"loss": 0.818468994140625,
"step": 9500
},
{
"epoch": 0.24871338651973446,
"grad_norm": 0.09459158033132553,
"learning_rate": 0.00041740596921855956,
"loss": 0.815874755859375,
"step": 9750
},
{
"epoch": 0.2550906528407533,
"grad_norm": 0.0948140025138855,
"learning_rate": 0.0004138629857430345,
"loss": 0.8106996459960938,
"step": 10000
},
{
"epoch": 0.2614679191617721,
"grad_norm": 0.09076733887195587,
"learning_rate": 0.0004103200022675094,
"loss": 0.8120379028320313,
"step": 10250
},
{
"epoch": 0.26784518548279096,
"grad_norm": 0.0840587168931961,
"learning_rate": 0.0004067770187919844,
"loss": 0.8110909423828125,
"step": 10500
},
{
"epoch": 0.2742224518038098,
"grad_norm": 0.08920229971408844,
"learning_rate": 0.0004032340353164593,
"loss": 0.80626806640625,
"step": 10750
},
{
"epoch": 0.2805997181248286,
"grad_norm": 0.0837017148733139,
"learning_rate": 0.00039969105184093425,
"loss": 0.8057803344726563,
"step": 11000
},
{
"epoch": 0.28697698444584746,
"grad_norm": 0.08721166849136353,
"learning_rate": 0.0003961480683654092,
"loss": 0.8038002319335937,
"step": 11250
},
{
"epoch": 0.2933542507668663,
"grad_norm": 0.0881330668926239,
"learning_rate": 0.00039260508488988405,
"loss": 0.801577880859375,
"step": 11500
},
{
"epoch": 0.2997315170878851,
"grad_norm": 0.09358793497085571,
"learning_rate": 0.000389062101414359,
"loss": 0.7983843994140625,
"step": 11750
},
{
"epoch": 0.30610878340890396,
"grad_norm": 0.08768702298402786,
"learning_rate": 0.0003855191179388339,
"loss": 0.7974132080078125,
"step": 12000
},
{
"epoch": 0.3124860497299228,
"grad_norm": 0.0864625945687294,
"learning_rate": 0.0003819761344633089,
"loss": 0.7967001342773438,
"step": 12250
},
{
"epoch": 0.3188633160509416,
"grad_norm": 0.08685760200023651,
"learning_rate": 0.0003784331509877838,
"loss": 0.7927449340820313,
"step": 12500
},
{
"epoch": 0.32524058237196046,
"grad_norm": 0.08900053054094315,
"learning_rate": 0.00037489016751225874,
"loss": 0.7911568603515625,
"step": 12750
},
{
"epoch": 0.3316178486929793,
"grad_norm": 0.08307567983865738,
"learning_rate": 0.00037134718403673367,
"loss": 0.7934022216796875,
"step": 13000
},
{
"epoch": 0.3379951150139981,
"grad_norm": 0.08461463451385498,
"learning_rate": 0.0003678042005612086,
"loss": 0.79057421875,
"step": 13250
},
{
"epoch": 0.3443723813350169,
"grad_norm": 0.08529062569141388,
"learning_rate": 0.00036426121708568347,
"loss": 0.7871109008789062,
"step": 13500
},
{
"epoch": 0.3507496476560358,
"grad_norm": 0.0872483178973198,
"learning_rate": 0.00036071823361015845,
"loss": 0.7870155029296875,
"step": 13750
},
{
"epoch": 0.3571269139770546,
"grad_norm": 0.0882478877902031,
"learning_rate": 0.0003571752501346334,
"loss": 0.78544140625,
"step": 14000
},
{
"epoch": 0.3635041802980734,
"grad_norm": 0.091294065117836,
"learning_rate": 0.0003536322666591083,
"loss": 0.7866491088867188,
"step": 14250
},
{
"epoch": 0.3698814466190923,
"grad_norm": 0.08572017401456833,
"learning_rate": 0.00035008928318358323,
"loss": 0.782495849609375,
"step": 14500
},
{
"epoch": 0.3762587129401111,
"grad_norm": 0.08324530720710754,
"learning_rate": 0.00034654629970805816,
"loss": 0.7797513427734375,
"step": 14750
},
{
"epoch": 0.3826359792611299,
"grad_norm": 0.08846472948789597,
"learning_rate": 0.0003430033162325331,
"loss": 0.7802353515625,
"step": 15000
},
{
"epoch": 0.3890132455821488,
"grad_norm": 0.07917249202728271,
"learning_rate": 0.00033946033275700807,
"loss": 0.7790340576171875,
"step": 15250
},
{
"epoch": 0.3953905119031676,
"grad_norm": 0.08665470778942108,
"learning_rate": 0.00033591734928148294,
"loss": 0.77759619140625,
"step": 15500
},
{
"epoch": 0.4017677782241864,
"grad_norm": 0.08258507400751114,
"learning_rate": 0.00033237436580595787,
"loss": 0.7763189086914063,
"step": 15750
},
{
"epoch": 0.4081450445452052,
"grad_norm": 0.08015473932027817,
"learning_rate": 0.0003288313823304328,
"loss": 0.7767020263671875,
"step": 16000
},
{
"epoch": 0.4145223108662241,
"grad_norm": 0.08080420643091202,
"learning_rate": 0.0003252883988549077,
"loss": 0.7747213134765625,
"step": 16250
},
{
"epoch": 0.4208995771872429,
"grad_norm": 0.08612637221813202,
"learning_rate": 0.00032174541537938265,
"loss": 0.7734822998046875,
"step": 16500
},
{
"epoch": 0.42727684350826173,
"grad_norm": 0.07643167674541473,
"learning_rate": 0.00031820243190385764,
"loss": 0.7744876708984375,
"step": 16750
},
{
"epoch": 0.4336541098292806,
"grad_norm": 0.08003860712051392,
"learning_rate": 0.00031465944842833256,
"loss": 0.770651611328125,
"step": 17000
},
{
"epoch": 0.4400313761502994,
"grad_norm": 0.07683893293142319,
"learning_rate": 0.0003111164649528075,
"loss": 0.770556640625,
"step": 17250
},
{
"epoch": 0.44640864247131823,
"grad_norm": 0.08421041071414948,
"learning_rate": 0.00030757348147728236,
"loss": 0.7692486572265625,
"step": 17500
},
{
"epoch": 0.4527859087923371,
"grad_norm": 0.08235982805490494,
"learning_rate": 0.0003040304980017573,
"loss": 0.7692238159179687,
"step": 17750
},
{
"epoch": 0.4591631751133559,
"grad_norm": 0.08046766370534897,
"learning_rate": 0.0003004875145262322,
"loss": 0.769811767578125,
"step": 18000
},
{
"epoch": 0.46554044143437473,
"grad_norm": 0.08188624680042267,
"learning_rate": 0.0002969445310507072,
"loss": 0.7680709228515625,
"step": 18250
},
{
"epoch": 0.4719177077553936,
"grad_norm": 0.0804840475320816,
"learning_rate": 0.00029340154757518213,
"loss": 0.7635980224609376,
"step": 18500
},
{
"epoch": 0.4782949740764124,
"grad_norm": 0.08123503625392914,
"learning_rate": 0.00028985856409965706,
"loss": 0.7640174560546875,
"step": 18750
},
{
"epoch": 0.48467224039743123,
"grad_norm": 0.08404447138309479,
"learning_rate": 0.000286315580624132,
"loss": 0.7631983642578125,
"step": 19000
},
{
"epoch": 0.49104950671845005,
"grad_norm": 0.09293997287750244,
"learning_rate": 0.0002827725971486069,
"loss": 0.7630362548828125,
"step": 19250
},
{
"epoch": 0.4974267730394689,
"grad_norm": 0.08382653445005417,
"learning_rate": 0.00027922961367308184,
"loss": 0.76153271484375,
"step": 19500
},
{
"epoch": 0.5038040393604878,
"grad_norm": 0.07555802911520004,
"learning_rate": 0.00027568663019755677,
"loss": 0.7605255737304687,
"step": 19750
},
{
"epoch": 0.5101813056815065,
"grad_norm": 0.08228955417871475,
"learning_rate": 0.0002721436467220317,
"loss": 0.7614573974609375,
"step": 20000
},
{
"epoch": 0.5165585720025254,
"grad_norm": 0.08145914226770401,
"learning_rate": 0.0002686006632465066,
"loss": 0.75997900390625,
"step": 20250
},
{
"epoch": 0.5229358383235442,
"grad_norm": 0.09308425337076187,
"learning_rate": 0.00026505767977098155,
"loss": 0.759119140625,
"step": 20500
},
{
"epoch": 0.529313104644563,
"grad_norm": 0.08138495683670044,
"learning_rate": 0.0002615146962954565,
"loss": 0.7591130981445312,
"step": 20750
},
{
"epoch": 0.5356903709655819,
"grad_norm": 0.08346766978502274,
"learning_rate": 0.0002579717128199314,
"loss": 0.757355224609375,
"step": 21000
},
{
"epoch": 0.5420676372866007,
"grad_norm": 0.077271968126297,
"learning_rate": 0.0002544287293444064,
"loss": 0.7534990234375,
"step": 21250
},
{
"epoch": 0.5484449036076195,
"grad_norm": 0.07977950572967529,
"learning_rate": 0.0002508857458688813,
"loss": 0.75655029296875,
"step": 21500
},
{
"epoch": 0.5548221699286384,
"grad_norm": 0.08347009867429733,
"learning_rate": 0.0002473427623933562,
"loss": 0.75475537109375,
"step": 21750
},
{
"epoch": 0.5611994362496572,
"grad_norm": 0.07832519710063934,
"learning_rate": 0.00024379977891783114,
"loss": 0.7558754272460938,
"step": 22000
},
{
"epoch": 0.567576702570676,
"grad_norm": 0.08469550311565399,
"learning_rate": 0.00024025679544230604,
"loss": 0.7546185913085938,
"step": 22250
},
{
"epoch": 0.5739539688916949,
"grad_norm": 0.08166459202766418,
"learning_rate": 0.000236713811966781,
"loss": 0.7514425659179688,
"step": 22500
},
{
"epoch": 0.5803312352127137,
"grad_norm": 0.08695302903652191,
"learning_rate": 0.00023317082849125592,
"loss": 0.75103564453125,
"step": 22750
},
{
"epoch": 0.5867085015337326,
"grad_norm": 0.09806793183088303,
"learning_rate": 0.00022962784501573085,
"loss": 0.752026123046875,
"step": 23000
},
{
"epoch": 0.5930857678547514,
"grad_norm": 0.08246204257011414,
"learning_rate": 0.00022608486154020578,
"loss": 0.7497666015625,
"step": 23250
},
{
"epoch": 0.5994630341757702,
"grad_norm": 0.08100980520248413,
"learning_rate": 0.0002225418780646807,
"loss": 0.749739501953125,
"step": 23500
},
{
"epoch": 0.605840300496789,
"grad_norm": 0.07755894213914871,
"learning_rate": 0.00021899889458915564,
"loss": 0.7490238037109375,
"step": 23750
},
{
"epoch": 0.6122175668178079,
"grad_norm": 0.08628904074430466,
"learning_rate": 0.0002154559111136306,
"loss": 0.7485498046875,
"step": 24000
},
{
"epoch": 0.6185948331388267,
"grad_norm": 0.08315210789442062,
"learning_rate": 0.0002119129276381055,
"loss": 0.7472803344726563,
"step": 24250
},
{
"epoch": 0.6249720994598456,
"grad_norm": 0.08354892581701279,
"learning_rate": 0.00020836994416258042,
"loss": 0.7469033203125,
"step": 24500
},
{
"epoch": 0.6313493657808644,
"grad_norm": 0.08090732991695404,
"learning_rate": 0.00020482696068705537,
"loss": 0.7483990478515625,
"step": 24750
},
{
"epoch": 0.6377266321018832,
"grad_norm": 0.08247523009777069,
"learning_rate": 0.0002012839772115303,
"loss": 0.7441016845703124,
"step": 25000
},
{
"epoch": 0.644103898422902,
"grad_norm": 0.07630934566259384,
"learning_rate": 0.0001977409937360052,
"loss": 0.741853515625,
"step": 25250
},
{
"epoch": 0.6504811647439209,
"grad_norm": 0.07914166897535324,
"learning_rate": 0.00019419801026048016,
"loss": 0.74271875,
"step": 25500
},
{
"epoch": 0.6568584310649397,
"grad_norm": 0.08192494511604309,
"learning_rate": 0.00019065502678495508,
"loss": 0.7423253784179688,
"step": 25750
},
{
"epoch": 0.6632356973859586,
"grad_norm": 0.08241453766822815,
"learning_rate": 0.00018711204330943,
"loss": 0.742218994140625,
"step": 26000
},
{
"epoch": 0.6696129637069773,
"grad_norm": 0.07896443456411362,
"learning_rate": 0.00018356905983390494,
"loss": 0.738341064453125,
"step": 26250
},
{
"epoch": 0.6759902300279962,
"grad_norm": 0.07580381631851196,
"learning_rate": 0.00018002607635837987,
"loss": 0.7400193481445313,
"step": 26500
},
{
"epoch": 0.6823674963490151,
"grad_norm": 0.07701326906681061,
"learning_rate": 0.0001764830928828548,
"loss": 0.7382138671875,
"step": 26750
},
{
"epoch": 0.6887447626700338,
"grad_norm": 0.07780223339796066,
"learning_rate": 0.00017294010940732975,
"loss": 0.7394266967773437,
"step": 27000
},
{
"epoch": 0.6951220289910527,
"grad_norm": 0.08848556876182556,
"learning_rate": 0.00016939712593180465,
"loss": 0.7408846435546875,
"step": 27250
},
{
"epoch": 0.7014992953120716,
"grad_norm": 0.0774645283818245,
"learning_rate": 0.00016585414245627958,
"loss": 0.7371785888671875,
"step": 27500
},
{
"epoch": 0.7078765616330903,
"grad_norm": 0.08030521869659424,
"learning_rate": 0.00016231115898075453,
"loss": 0.7363970947265625,
"step": 27750
},
{
"epoch": 0.7142538279541092,
"grad_norm": 0.08429484069347382,
"learning_rate": 0.00015876817550522946,
"loss": 0.7369862670898437,
"step": 28000
},
{
"epoch": 0.7206310942751281,
"grad_norm": 0.07752969861030579,
"learning_rate": 0.00015522519202970436,
"loss": 0.735985595703125,
"step": 28250
},
{
"epoch": 0.7270083605961468,
"grad_norm": 0.08850430697202682,
"learning_rate": 0.0001516822085541793,
"loss": 0.7350040893554688,
"step": 28500
},
{
"epoch": 0.7333856269171657,
"grad_norm": 0.07794130593538284,
"learning_rate": 0.00014813922507865424,
"loss": 0.7345814208984375,
"step": 28750
},
{
"epoch": 0.7397628932381846,
"grad_norm": 0.07695607095956802,
"learning_rate": 0.00014459624160312917,
"loss": 0.731819091796875,
"step": 29000
},
{
"epoch": 0.7461401595592033,
"grad_norm": 0.07952513545751572,
"learning_rate": 0.0001410532581276041,
"loss": 0.7328406982421874,
"step": 29250
},
{
"epoch": 0.7525174258802222,
"grad_norm": 0.08443118631839752,
"learning_rate": 0.00013751027465207902,
"loss": 0.7327833862304688,
"step": 29500
},
{
"epoch": 0.7588946922012411,
"grad_norm": 0.07715488970279694,
"learning_rate": 0.00013396729117655395,
"loss": 0.7313681030273438,
"step": 29750
},
{
"epoch": 0.7652719585222598,
"grad_norm": 0.08632020652294159,
"learning_rate": 0.0001304243077010289,
"loss": 0.7293426513671875,
"step": 30000
},
{
"epoch": 0.7716492248432787,
"grad_norm": 0.08144113421440125,
"learning_rate": 0.0001268813242255038,
"loss": 0.729284912109375,
"step": 30250
},
{
"epoch": 0.7780264911642976,
"grad_norm": 0.08117876201868057,
"learning_rate": 0.00012333834074997873,
"loss": 0.7302175903320313,
"step": 30500
},
{
"epoch": 0.7844037574853163,
"grad_norm": 0.08066125214099884,
"learning_rate": 0.00011979535727445367,
"loss": 0.7301824951171875,
"step": 30750
},
{
"epoch": 0.7907810238063352,
"grad_norm": 0.08321163803339005,
"learning_rate": 0.0001162523737989286,
"loss": 0.726185302734375,
"step": 31000
},
{
"epoch": 0.7971582901273541,
"grad_norm": 0.07980816811323166,
"learning_rate": 0.00011270939032340353,
"loss": 0.7280133056640625,
"step": 31250
},
{
"epoch": 0.8035355564483728,
"grad_norm": 0.0806281566619873,
"learning_rate": 0.00010916640684787846,
"loss": 0.7268768310546875,
"step": 31500
},
{
"epoch": 0.8099128227693917,
"grad_norm": 0.08116059750318527,
"learning_rate": 0.0001056234233723534,
"loss": 0.7262345581054688,
"step": 31750
},
{
"epoch": 0.8162900890904105,
"grad_norm": 0.07698776572942734,
"learning_rate": 0.00010208043989682831,
"loss": 0.7262030029296875,
"step": 32000
},
{
"epoch": 0.8226673554114293,
"grad_norm": 0.08539737015962601,
"learning_rate": 9.853745642130325e-05,
"loss": 0.7263545532226563,
"step": 32250
},
{
"epoch": 0.8290446217324482,
"grad_norm": 0.08154796808958054,
"learning_rate": 9.499447294577818e-05,
"loss": 0.7248871459960937,
"step": 32500
},
{
"epoch": 0.835421888053467,
"grad_norm": 0.08672577142715454,
"learning_rate": 9.145148947025311e-05,
"loss": 0.7239915771484375,
"step": 32750
},
{
"epoch": 0.8417991543744858,
"grad_norm": 0.08093911409378052,
"learning_rate": 8.790850599472804e-05,
"loss": 0.723066162109375,
"step": 33000
},
{
"epoch": 0.8481764206955047,
"grad_norm": 0.07659825682640076,
"learning_rate": 8.436552251920298e-05,
"loss": 0.72345263671875,
"step": 33250
},
{
"epoch": 0.8545536870165235,
"grad_norm": 0.08022043108940125,
"learning_rate": 8.082253904367789e-05,
"loss": 0.72362255859375,
"step": 33500
},
{
"epoch": 0.8609309533375423,
"grad_norm": 0.08199780434370041,
"learning_rate": 7.727955556815283e-05,
"loss": 0.7216689453125,
"step": 33750
},
{
"epoch": 0.8673082196585612,
"grad_norm": 0.07852373272180557,
"learning_rate": 7.373657209262777e-05,
"loss": 0.7197748413085937,
"step": 34000
},
{
"epoch": 0.87368548597958,
"grad_norm": 0.07824918627738953,
"learning_rate": 7.019358861710269e-05,
"loss": 0.7181076049804688,
"step": 34250
},
{
"epoch": 0.8800627523005988,
"grad_norm": 0.07989538460969925,
"learning_rate": 6.665060514157763e-05,
"loss": 0.7219158325195313,
"step": 34500
},
{
"epoch": 0.8864400186216177,
"grad_norm": 0.07987392693758011,
"learning_rate": 6.310762166605256e-05,
"loss": 0.720794189453125,
"step": 34750
},
{
"epoch": 0.8928172849426365,
"grad_norm": 0.08085554093122482,
"learning_rate": 5.956463819052748e-05,
"loss": 0.7185697021484375,
"step": 35000
},
{
"epoch": 0.8991945512636553,
"grad_norm": 0.0825972706079483,
"learning_rate": 5.602165471500241e-05,
"loss": 0.7219225463867187,
"step": 35250
},
{
"epoch": 0.9055718175846742,
"grad_norm": 0.08420496433973312,
"learning_rate": 5.247867123947734e-05,
"loss": 0.7178782348632813,
"step": 35500
},
{
"epoch": 0.911949083905693,
"grad_norm": 0.07988401502370834,
"learning_rate": 4.893568776395227e-05,
"loss": 0.7176174926757812,
"step": 35750
},
{
"epoch": 0.9183263502267118,
"grad_norm": 0.08001978695392609,
"learning_rate": 4.53927042884272e-05,
"loss": 0.7183565673828125,
"step": 36000
},
{
"epoch": 0.9247036165477307,
"grad_norm": 0.07920462638139725,
"learning_rate": 4.184972081290213e-05,
"loss": 0.7176240234375,
"step": 36250
},
{
"epoch": 0.9310808828687495,
"grad_norm": 0.08121761679649353,
"learning_rate": 3.8306737337377056e-05,
"loss": 0.71547265625,
"step": 36500
},
{
"epoch": 0.9374581491897683,
"grad_norm": 0.07785205543041229,
"learning_rate": 3.476375386185199e-05,
"loss": 0.7175172729492187,
"step": 36750
},
{
"epoch": 0.9438354155107872,
"grad_norm": 0.07759752124547958,
"learning_rate": 3.122077038632692e-05,
"loss": 0.7157890625,
"step": 37000
},
{
"epoch": 0.950212681831806,
"grad_norm": 0.08237745612859726,
"learning_rate": 2.7677786910801846e-05,
"loss": 0.7169793701171875,
"step": 37250
},
{
"epoch": 0.9565899481528248,
"grad_norm": 0.07707587629556656,
"learning_rate": 2.4134803435276777e-05,
"loss": 0.715412109375,
"step": 37500
},
{
"epoch": 0.9629672144738436,
"grad_norm": 0.07932782173156738,
"learning_rate": 2.059181995975171e-05,
"loss": 0.71594677734375,
"step": 37750
},
{
"epoch": 0.9693444807948625,
"grad_norm": 0.08169736713171005,
"learning_rate": 1.704883648422664e-05,
"loss": 0.7143995361328125,
"step": 38000
},
{
"epoch": 0.9757217471158813,
"grad_norm": 0.08160339295864105,
"learning_rate": 1.3505853008701568e-05,
"loss": 0.7135118408203125,
"step": 38250
},
{
"epoch": 0.9820990134369001,
"grad_norm": 0.08456628769636154,
"learning_rate": 9.962869533176497e-06,
"loss": 0.714119384765625,
"step": 38500
},
{
"epoch": 0.988476279757919,
"grad_norm": 0.07943534851074219,
"learning_rate": 6.419886057651427e-06,
"loss": 0.7134039306640625,
"step": 38750
},
{
"epoch": 0.9948535460789378,
"grad_norm": 0.0769161656498909,
"learning_rate": 2.8769025821263568e-06,
"loss": 0.71384814453125,
"step": 39000
},
{
"epoch": 1.0,
"step": 39202,
"total_flos": 7.006185158986433e+18,
"train_loss": 0.8243957067240222,
"train_runtime": 51861.4856,
"train_samples_per_second": 48.377,
"train_steps_per_second": 0.756
}
],
"logging_steps": 250,
"max_steps": 39202,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 3921,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 7.006185158986433e+18,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}