Files
AetherCuTF-Ko-2-shared-Seed…/trainer_state.json
ModelHub XC b083bf55e7 初始化项目,由ModelHub XC社区提供模型
Model: AetherModels/AetherCuTF-Ko-2-shared-Seed1234
Source: Original Platform
2026-07-13 11:45:11 +08:00

1136 lines
30 KiB
JSON

{
"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": 3.069216251373291,
"learning_rate": 3.175210405508799e-05,
"loss": 3.66965771484375,
"step": 250
},
{
"epoch": 0.012754532642037663,
"grad_norm": 3.3416852951049805,
"learning_rate": 6.363172660035706e-05,
"loss": 2.034814208984375,
"step": 500
},
{
"epoch": 0.019131798963056498,
"grad_norm": 2.1286585330963135,
"learning_rate": 9.551134914562612e-05,
"loss": 1.61390625,
"step": 750
},
{
"epoch": 0.025509065284075327,
"grad_norm": 1.109268307685852,
"learning_rate": 0.00012739097169089516,
"loss": 1.44977294921875,
"step": 1000
},
{
"epoch": 0.03188633160509416,
"grad_norm": 0.7057593464851379,
"learning_rate": 0.00015927059423616425,
"loss": 1.3364879150390625,
"step": 1250
},
{
"epoch": 0.038263597926112995,
"grad_norm": 0.47604697942733765,
"learning_rate": 0.0001911502167814333,
"loss": 1.22646484375,
"step": 1500
},
{
"epoch": 0.044640864247131824,
"grad_norm": 0.3333810567855835,
"learning_rate": 0.00022302983932670237,
"loss": 1.1279931640625,
"step": 1750
},
{
"epoch": 0.05101813056815065,
"grad_norm": 0.2781062722206116,
"learning_rate": 0.00025490946187197144,
"loss": 1.0727122802734375,
"step": 2000
},
{
"epoch": 0.05739539688916949,
"grad_norm": 0.26381585001945496,
"learning_rate": 0.00028678908441724047,
"loss": 1.03530322265625,
"step": 2250
},
{
"epoch": 0.06377266321018832,
"grad_norm": 0.20349043607711792,
"learning_rate": 0.00031866870696250956,
"loss": 1.011820556640625,
"step": 2500
},
{
"epoch": 0.07014992953120715,
"grad_norm": 0.20157715678215027,
"learning_rate": 0.00035054832950777865,
"loss": 0.9910045166015625,
"step": 2750
},
{
"epoch": 0.07652719585222599,
"grad_norm": 0.16963635385036469,
"learning_rate": 0.0003824279520530477,
"loss": 0.9719039306640626,
"step": 3000
},
{
"epoch": 0.08290446217324482,
"grad_norm": 0.14934051036834717,
"learning_rate": 0.0004143075745983167,
"loss": 0.9600828857421875,
"step": 3250
},
{
"epoch": 0.08928172849426365,
"grad_norm": 0.14919009804725647,
"learning_rate": 0.00044618719714358587,
"loss": 0.9467827758789062,
"step": 3500
},
{
"epoch": 0.09565899481528248,
"grad_norm": 0.1348155289888382,
"learning_rate": 0.0004780668196888549,
"loss": 0.9338873291015625,
"step": 3750
},
{
"epoch": 0.1020362611363013,
"grad_norm": 0.13575714826583862,
"learning_rate": 0.0004988945891556362,
"loss": 0.927226806640625,
"step": 4000
},
{
"epoch": 0.10841352745732015,
"grad_norm": 0.12741005420684814,
"learning_rate": 0.0004953516056801111,
"loss": 0.9160831298828125,
"step": 4250
},
{
"epoch": 0.11479079377833898,
"grad_norm": 0.116083063185215,
"learning_rate": 0.0004918086222045861,
"loss": 0.909201904296875,
"step": 4500
},
{
"epoch": 0.12116806009935781,
"grad_norm": 0.1180315837264061,
"learning_rate": 0.000488265638729061,
"loss": 0.897420654296875,
"step": 4750
},
{
"epoch": 0.12754532642037664,
"grad_norm": 0.12467160820960999,
"learning_rate": 0.00048472265525353593,
"loss": 0.8899200439453125,
"step": 5000
},
{
"epoch": 0.13392259274139548,
"grad_norm": 0.11566604673862457,
"learning_rate": 0.00048117967177801086,
"loss": 0.8843126220703125,
"step": 5250
},
{
"epoch": 0.1402998590624143,
"grad_norm": 0.10344680398702621,
"learning_rate": 0.00047763668830248573,
"loss": 0.8766544189453125,
"step": 5500
},
{
"epoch": 0.14667712538343314,
"grad_norm": 0.09983552992343903,
"learning_rate": 0.00047409370482696066,
"loss": 0.8693883056640626,
"step": 5750
},
{
"epoch": 0.15305439170445198,
"grad_norm": 0.10290490090847015,
"learning_rate": 0.00047055072135143564,
"loss": 0.866027587890625,
"step": 6000
},
{
"epoch": 0.1594316580254708,
"grad_norm": 0.10045526921749115,
"learning_rate": 0.00046700773787591057,
"loss": 0.8622779541015625,
"step": 6250
},
{
"epoch": 0.16580892434648964,
"grad_norm": 0.09836139529943466,
"learning_rate": 0.0004634647544003855,
"loss": 0.8554696044921875,
"step": 6500
},
{
"epoch": 0.17218619066750845,
"grad_norm": 0.09789058566093445,
"learning_rate": 0.0004599217709248604,
"loss": 0.8513157348632813,
"step": 6750
},
{
"epoch": 0.1785634569885273,
"grad_norm": 0.09722394496202469,
"learning_rate": 0.00045637878744933535,
"loss": 0.8466351318359375,
"step": 7000
},
{
"epoch": 0.18494072330954614,
"grad_norm": 0.09477878361940384,
"learning_rate": 0.0004528358039738103,
"loss": 0.8434057006835938,
"step": 7250
},
{
"epoch": 0.19131798963056496,
"grad_norm": 0.10017236322164536,
"learning_rate": 0.0004492928204982852,
"loss": 0.8430394287109375,
"step": 7500
},
{
"epoch": 0.1976952559515838,
"grad_norm": 0.09258312731981277,
"learning_rate": 0.00044574983702276013,
"loss": 0.83903076171875,
"step": 7750
},
{
"epoch": 0.2040725222726026,
"grad_norm": 0.09747657179832458,
"learning_rate": 0.00044220685354723506,
"loss": 0.8325919799804687,
"step": 8000
},
{
"epoch": 0.21044978859362146,
"grad_norm": 0.09473184496164322,
"learning_rate": 0.00043866387007171,
"loss": 0.8321695556640625,
"step": 8250
},
{
"epoch": 0.2168270549146403,
"grad_norm": 0.09705399721860886,
"learning_rate": 0.0004351208865961849,
"loss": 0.8288480224609375,
"step": 8500
},
{
"epoch": 0.22320432123565911,
"grad_norm": 0.09528972208499908,
"learning_rate": 0.00043157790312065985,
"loss": 0.8265286254882812,
"step": 8750
},
{
"epoch": 0.22958158755667796,
"grad_norm": 0.0975126326084137,
"learning_rate": 0.0004280349196451348,
"loss": 0.825369140625,
"step": 9000
},
{
"epoch": 0.2359588538776968,
"grad_norm": 0.08992481231689453,
"learning_rate": 0.00042449193616960975,
"loss": 0.8229822998046875,
"step": 9250
},
{
"epoch": 0.24233612019871562,
"grad_norm": 0.087989941239357,
"learning_rate": 0.00042094895269408463,
"loss": 0.8201241455078125,
"step": 9500
},
{
"epoch": 0.24871338651973446,
"grad_norm": 0.09176884591579437,
"learning_rate": 0.00041740596921855956,
"loss": 0.8174908447265625,
"step": 9750
},
{
"epoch": 0.2550906528407533,
"grad_norm": 0.09006844460964203,
"learning_rate": 0.0004138629857430345,
"loss": 0.8124500732421875,
"step": 10000
},
{
"epoch": 0.2614679191617721,
"grad_norm": 0.09052397310733795,
"learning_rate": 0.0004103200022675094,
"loss": 0.8137666015625,
"step": 10250
},
{
"epoch": 0.26784518548279096,
"grad_norm": 0.08030358701944351,
"learning_rate": 0.0004067770187919844,
"loss": 0.8128126831054687,
"step": 10500
},
{
"epoch": 0.2742224518038098,
"grad_norm": 0.0884530320763588,
"learning_rate": 0.0004032340353164593,
"loss": 0.8079989624023437,
"step": 10750
},
{
"epoch": 0.2805997181248286,
"grad_norm": 0.08455190062522888,
"learning_rate": 0.00039969105184093425,
"loss": 0.80756201171875,
"step": 11000
},
{
"epoch": 0.28697698444584746,
"grad_norm": 0.09025362879037857,
"learning_rate": 0.0003961480683654092,
"loss": 0.8056405029296875,
"step": 11250
},
{
"epoch": 0.2933542507668663,
"grad_norm": 0.0882817953824997,
"learning_rate": 0.00039260508488988405,
"loss": 0.8034854736328125,
"step": 11500
},
{
"epoch": 0.2997315170878851,
"grad_norm": 0.08535934239625931,
"learning_rate": 0.000389062101414359,
"loss": 0.8002091064453125,
"step": 11750
},
{
"epoch": 0.30610878340890396,
"grad_norm": 0.08830045908689499,
"learning_rate": 0.0003855191179388339,
"loss": 0.7992052612304688,
"step": 12000
},
{
"epoch": 0.3124860497299228,
"grad_norm": 0.08184854686260223,
"learning_rate": 0.0003819761344633089,
"loss": 0.7985946044921876,
"step": 12250
},
{
"epoch": 0.3188633160509416,
"grad_norm": 0.08994520455598831,
"learning_rate": 0.0003784331509877838,
"loss": 0.794483154296875,
"step": 12500
},
{
"epoch": 0.32524058237196046,
"grad_norm": 0.08402872830629349,
"learning_rate": 0.00037489016751225874,
"loss": 0.793040283203125,
"step": 12750
},
{
"epoch": 0.3316178486929793,
"grad_norm": 0.08523915708065033,
"learning_rate": 0.00037134718403673367,
"loss": 0.7952125244140625,
"step": 13000
},
{
"epoch": 0.3379951150139981,
"grad_norm": 0.08320575952529907,
"learning_rate": 0.0003678042005612086,
"loss": 0.7924825439453125,
"step": 13250
},
{
"epoch": 0.3443723813350169,
"grad_norm": 0.07906585186719894,
"learning_rate": 0.00036426121708568347,
"loss": 0.7891689453125,
"step": 13500
},
{
"epoch": 0.3507496476560358,
"grad_norm": 0.08302329480648041,
"learning_rate": 0.00036071823361015845,
"loss": 0.78904248046875,
"step": 13750
},
{
"epoch": 0.3571269139770546,
"grad_norm": 0.08724895864725113,
"learning_rate": 0.0003571752501346334,
"loss": 0.787458740234375,
"step": 14000
},
{
"epoch": 0.3635041802980734,
"grad_norm": 0.08670660108327866,
"learning_rate": 0.0003536322666591083,
"loss": 0.7885794677734375,
"step": 14250
},
{
"epoch": 0.3698814466190923,
"grad_norm": 0.08605735749006271,
"learning_rate": 0.00035008928318358323,
"loss": 0.7844781494140625,
"step": 14500
},
{
"epoch": 0.3762587129401111,
"grad_norm": 0.08505359292030334,
"learning_rate": 0.00034654629970805816,
"loss": 0.7817347412109374,
"step": 14750
},
{
"epoch": 0.3826359792611299,
"grad_norm": 0.0848836600780487,
"learning_rate": 0.0003430033162325331,
"loss": 0.7822146606445313,
"step": 15000
},
{
"epoch": 0.3890132455821488,
"grad_norm": 0.07499703764915466,
"learning_rate": 0.00033946033275700807,
"loss": 0.7810926513671875,
"step": 15250
},
{
"epoch": 0.3953905119031676,
"grad_norm": 0.08446674793958664,
"learning_rate": 0.00033591734928148294,
"loss": 0.7797041625976563,
"step": 15500
},
{
"epoch": 0.4017677782241864,
"grad_norm": 0.08182055503129959,
"learning_rate": 0.00033237436580595787,
"loss": 0.7783648681640625,
"step": 15750
},
{
"epoch": 0.4081450445452052,
"grad_norm": 0.07750219851732254,
"learning_rate": 0.0003288313823304328,
"loss": 0.778696533203125,
"step": 16000
},
{
"epoch": 0.4145223108662241,
"grad_norm": 0.07737597823143005,
"learning_rate": 0.0003252883988549077,
"loss": 0.7767322998046875,
"step": 16250
},
{
"epoch": 0.4208995771872429,
"grad_norm": 0.08738387376070023,
"learning_rate": 0.00032174541537938265,
"loss": 0.7756185302734375,
"step": 16500
},
{
"epoch": 0.42727684350826173,
"grad_norm": 0.07566338032484055,
"learning_rate": 0.00031820243190385764,
"loss": 0.7765888061523437,
"step": 16750
},
{
"epoch": 0.4336541098292806,
"grad_norm": 0.08135856688022614,
"learning_rate": 0.00031465944842833256,
"loss": 0.772817138671875,
"step": 17000
},
{
"epoch": 0.4400313761502994,
"grad_norm": 0.07488487660884857,
"learning_rate": 0.0003111164649528075,
"loss": 0.7726251831054688,
"step": 17250
},
{
"epoch": 0.44640864247131823,
"grad_norm": 0.08308962732553482,
"learning_rate": 0.00030757348147728236,
"loss": 0.7713480224609375,
"step": 17500
},
{
"epoch": 0.4527859087923371,
"grad_norm": 0.0840180441737175,
"learning_rate": 0.0003040304980017573,
"loss": 0.7713153686523437,
"step": 17750
},
{
"epoch": 0.4591631751133559,
"grad_norm": 0.08091706782579422,
"learning_rate": 0.0003004875145262322,
"loss": 0.771861572265625,
"step": 18000
},
{
"epoch": 0.46554044143437473,
"grad_norm": 0.07731328904628754,
"learning_rate": 0.0002969445310507072,
"loss": 0.7702142333984375,
"step": 18250
},
{
"epoch": 0.4719177077553936,
"grad_norm": 0.079083651304245,
"learning_rate": 0.00029340154757518213,
"loss": 0.7657806396484375,
"step": 18500
},
{
"epoch": 0.4782949740764124,
"grad_norm": 0.07866887748241425,
"learning_rate": 0.00028985856409965706,
"loss": 0.76620263671875,
"step": 18750
},
{
"epoch": 0.48467224039743123,
"grad_norm": 0.07663533836603165,
"learning_rate": 0.000286315580624132,
"loss": 0.7652430419921875,
"step": 19000
},
{
"epoch": 0.49104950671845005,
"grad_norm": 0.08650851994752884,
"learning_rate": 0.0002827725971486069,
"loss": 0.7652618408203125,
"step": 19250
},
{
"epoch": 0.4974267730394689,
"grad_norm": 0.07804276794195175,
"learning_rate": 0.00027922961367308184,
"loss": 0.7637146606445312,
"step": 19500
},
{
"epoch": 0.5038040393604878,
"grad_norm": 0.0775287076830864,
"learning_rate": 0.00027568663019755677,
"loss": 0.7627333984375,
"step": 19750
},
{
"epoch": 0.5101813056815065,
"grad_norm": 0.07739653438329697,
"learning_rate": 0.0002721436467220317,
"loss": 0.7636646728515625,
"step": 20000
},
{
"epoch": 0.5165585720025254,
"grad_norm": 0.07964754849672318,
"learning_rate": 0.0002686006632465066,
"loss": 0.7622118530273437,
"step": 20250
},
{
"epoch": 0.5229358383235442,
"grad_norm": 0.08351464569568634,
"learning_rate": 0.00026505767977098155,
"loss": 0.7613211059570313,
"step": 20500
},
{
"epoch": 0.529313104644563,
"grad_norm": 0.07561041414737701,
"learning_rate": 0.0002615146962954565,
"loss": 0.761345458984375,
"step": 20750
},
{
"epoch": 0.5356903709655819,
"grad_norm": 0.07760997861623764,
"learning_rate": 0.0002579717128199314,
"loss": 0.7596089477539063,
"step": 21000
},
{
"epoch": 0.5420676372866007,
"grad_norm": 0.07653113454580307,
"learning_rate": 0.0002544287293444064,
"loss": 0.7557307739257813,
"step": 21250
},
{
"epoch": 0.5484449036076195,
"grad_norm": 0.07422952353954315,
"learning_rate": 0.0002508857458688813,
"loss": 0.7588695678710937,
"step": 21500
},
{
"epoch": 0.5548221699286384,
"grad_norm": 0.07543788850307465,
"learning_rate": 0.0002473427623933562,
"loss": 0.757001708984375,
"step": 21750
},
{
"epoch": 0.5611994362496572,
"grad_norm": 0.08028650283813477,
"learning_rate": 0.00024379977891783114,
"loss": 0.7580865478515625,
"step": 22000
},
{
"epoch": 0.567576702570676,
"grad_norm": 0.07965358346700668,
"learning_rate": 0.00024025679544230604,
"loss": 0.7568743896484375,
"step": 22250
},
{
"epoch": 0.5739539688916949,
"grad_norm": 0.0798027515411377,
"learning_rate": 0.000236713811966781,
"loss": 0.7538323974609376,
"step": 22500
},
{
"epoch": 0.5803312352127137,
"grad_norm": 0.08517932146787643,
"learning_rate": 0.00023317082849125592,
"loss": 0.753489990234375,
"step": 22750
},
{
"epoch": 0.5867085015337326,
"grad_norm": 0.0870809406042099,
"learning_rate": 0.00022962784501573085,
"loss": 0.7543722534179688,
"step": 23000
},
{
"epoch": 0.5930857678547514,
"grad_norm": 0.07744799554347992,
"learning_rate": 0.00022608486154020578,
"loss": 0.7519418334960938,
"step": 23250
},
{
"epoch": 0.5994630341757702,
"grad_norm": 0.07942230254411697,
"learning_rate": 0.0002225418780646807,
"loss": 0.7520146484375,
"step": 23500
},
{
"epoch": 0.605840300496789,
"grad_norm": 0.07722324877977371,
"learning_rate": 0.00021899889458915564,
"loss": 0.7513837280273438,
"step": 23750
},
{
"epoch": 0.6122175668178079,
"grad_norm": 0.08005702495574951,
"learning_rate": 0.0002154559111136306,
"loss": 0.750864501953125,
"step": 24000
},
{
"epoch": 0.6185948331388267,
"grad_norm": 0.08771445602178574,
"learning_rate": 0.0002119129276381055,
"loss": 0.7495726928710937,
"step": 24250
},
{
"epoch": 0.6249720994598456,
"grad_norm": 0.07859943062067032,
"learning_rate": 0.00020836994416258042,
"loss": 0.7492670288085937,
"step": 24500
},
{
"epoch": 0.6313493657808644,
"grad_norm": 0.0817301794886589,
"learning_rate": 0.00020482696068705537,
"loss": 0.7507242431640625,
"step": 24750
},
{
"epoch": 0.6377266321018832,
"grad_norm": 0.08105339854955673,
"learning_rate": 0.0002012839772115303,
"loss": 0.7463834228515625,
"step": 25000
},
{
"epoch": 0.644103898422902,
"grad_norm": 0.07463200390338898,
"learning_rate": 0.0001977409937360052,
"loss": 0.7442008666992187,
"step": 25250
},
{
"epoch": 0.6504811647439209,
"grad_norm": 0.07825867086648941,
"learning_rate": 0.00019419801026048016,
"loss": 0.74505517578125,
"step": 25500
},
{
"epoch": 0.6568584310649397,
"grad_norm": 0.07815507054328918,
"learning_rate": 0.00019065502678495508,
"loss": 0.7447003784179688,
"step": 25750
},
{
"epoch": 0.6632356973859586,
"grad_norm": 0.08679299056529999,
"learning_rate": 0.00018711204330943,
"loss": 0.74454541015625,
"step": 26000
},
{
"epoch": 0.6696129637069773,
"grad_norm": 0.07593758404254913,
"learning_rate": 0.00018356905983390494,
"loss": 0.740745849609375,
"step": 26250
},
{
"epoch": 0.6759902300279962,
"grad_norm": 0.07361768186092377,
"learning_rate": 0.00018002607635837987,
"loss": 0.742380126953125,
"step": 26500
},
{
"epoch": 0.6823674963490151,
"grad_norm": 0.07657358795404434,
"learning_rate": 0.0001764830928828548,
"loss": 0.7406201171875,
"step": 26750
},
{
"epoch": 0.6887447626700338,
"grad_norm": 0.07750886678695679,
"learning_rate": 0.00017294010940732975,
"loss": 0.74180078125,
"step": 27000
},
{
"epoch": 0.6951220289910527,
"grad_norm": 0.08482435345649719,
"learning_rate": 0.00016939712593180465,
"loss": 0.743335205078125,
"step": 27250
},
{
"epoch": 0.7014992953120716,
"grad_norm": 0.07442943751811981,
"learning_rate": 0.00016585414245627958,
"loss": 0.739550537109375,
"step": 27500
},
{
"epoch": 0.7078765616330903,
"grad_norm": 0.07888231426477432,
"learning_rate": 0.00016231115898075453,
"loss": 0.738810791015625,
"step": 27750
},
{
"epoch": 0.7142538279541092,
"grad_norm": 0.08407917618751526,
"learning_rate": 0.00015876817550522946,
"loss": 0.7393695068359375,
"step": 28000
},
{
"epoch": 0.7206310942751281,
"grad_norm": 0.07510771602392197,
"learning_rate": 0.00015522519202970436,
"loss": 0.738358154296875,
"step": 28250
},
{
"epoch": 0.7270083605961468,
"grad_norm": 0.08540713787078857,
"learning_rate": 0.0001516822085541793,
"loss": 0.737420654296875,
"step": 28500
},
{
"epoch": 0.7333856269171657,
"grad_norm": 0.07714398205280304,
"learning_rate": 0.00014813922507865424,
"loss": 0.7370699462890625,
"step": 28750
},
{
"epoch": 0.7397628932381846,
"grad_norm": 0.07369086891412735,
"learning_rate": 0.00014459624160312917,
"loss": 0.7341400146484375,
"step": 29000
},
{
"epoch": 0.7461401595592033,
"grad_norm": 0.0806751474738121,
"learning_rate": 0.0001410532581276041,
"loss": 0.7352272338867187,
"step": 29250
},
{
"epoch": 0.7525174258802222,
"grad_norm": 0.08281698077917099,
"learning_rate": 0.00013751027465207902,
"loss": 0.7351591186523437,
"step": 29500
},
{
"epoch": 0.7588946922012411,
"grad_norm": 0.07720359414815903,
"learning_rate": 0.00013396729117655395,
"loss": 0.7339112548828125,
"step": 29750
},
{
"epoch": 0.7652719585222598,
"grad_norm": 0.08155587315559387,
"learning_rate": 0.0001304243077010289,
"loss": 0.7317435302734375,
"step": 30000
},
{
"epoch": 0.7716492248432787,
"grad_norm": 0.07980415970087051,
"learning_rate": 0.0001268813242255038,
"loss": 0.7317054443359375,
"step": 30250
},
{
"epoch": 0.7780264911642976,
"grad_norm": 0.07981084287166595,
"learning_rate": 0.00012333834074997873,
"loss": 0.73269287109375,
"step": 30500
},
{
"epoch": 0.7844037574853163,
"grad_norm": 0.080998994410038,
"learning_rate": 0.00011979535727445367,
"loss": 0.732618408203125,
"step": 30750
},
{
"epoch": 0.7907810238063352,
"grad_norm": 0.08089172095060349,
"learning_rate": 0.0001162523737989286,
"loss": 0.7286431884765625,
"step": 31000
},
{
"epoch": 0.7971582901273541,
"grad_norm": 0.07939846813678741,
"learning_rate": 0.00011270939032340353,
"loss": 0.7305592651367188,
"step": 31250
},
{
"epoch": 0.8035355564483728,
"grad_norm": 0.07798649370670319,
"learning_rate": 0.00010916640684787846,
"loss": 0.729310302734375,
"step": 31500
},
{
"epoch": 0.8099128227693917,
"grad_norm": 0.07753942161798477,
"learning_rate": 0.0001056234233723534,
"loss": 0.7285136108398438,
"step": 31750
},
{
"epoch": 0.8162900890904105,
"grad_norm": 0.07483146339654922,
"learning_rate": 0.00010208043989682831,
"loss": 0.7287723388671875,
"step": 32000
},
{
"epoch": 0.8226673554114293,
"grad_norm": 0.08278854191303253,
"learning_rate": 9.853745642130325e-05,
"loss": 0.7288785400390625,
"step": 32250
},
{
"epoch": 0.8290446217324482,
"grad_norm": 0.0812024474143982,
"learning_rate": 9.499447294577818e-05,
"loss": 0.7274485473632812,
"step": 32500
},
{
"epoch": 0.835421888053467,
"grad_norm": 0.081851527094841,
"learning_rate": 9.145148947025311e-05,
"loss": 0.7264241943359375,
"step": 32750
},
{
"epoch": 0.8417991543744858,
"grad_norm": 0.08262418955564499,
"learning_rate": 8.790850599472804e-05,
"loss": 0.7254639892578125,
"step": 33000
},
{
"epoch": 0.8481764206955047,
"grad_norm": 0.07723336666822433,
"learning_rate": 8.436552251920298e-05,
"loss": 0.7259024658203125,
"step": 33250
},
{
"epoch": 0.8545536870165235,
"grad_norm": 0.07814474403858185,
"learning_rate": 8.082253904367789e-05,
"loss": 0.7260317993164063,
"step": 33500
},
{
"epoch": 0.8609309533375423,
"grad_norm": 0.08336895704269409,
"learning_rate": 7.727955556815283e-05,
"loss": 0.724099365234375,
"step": 33750
},
{
"epoch": 0.8673082196585612,
"grad_norm": 0.07718467712402344,
"learning_rate": 7.373657209262777e-05,
"loss": 0.722183349609375,
"step": 34000
},
{
"epoch": 0.87368548597958,
"grad_norm": 0.07709118723869324,
"learning_rate": 7.019358861710269e-05,
"loss": 0.7206033935546875,
"step": 34250
},
{
"epoch": 0.8800627523005988,
"grad_norm": 0.07927864789962769,
"learning_rate": 6.665060514157763e-05,
"loss": 0.7243323974609375,
"step": 34500
},
{
"epoch": 0.8864400186216177,
"grad_norm": 0.0789353996515274,
"learning_rate": 6.310762166605256e-05,
"loss": 0.7233385009765625,
"step": 34750
},
{
"epoch": 0.8928172849426365,
"grad_norm": 0.07941902428865433,
"learning_rate": 5.956463819052748e-05,
"loss": 0.7211009521484375,
"step": 35000
},
{
"epoch": 0.8991945512636553,
"grad_norm": 0.08030779659748077,
"learning_rate": 5.602165471500241e-05,
"loss": 0.724384033203125,
"step": 35250
},
{
"epoch": 0.9055718175846742,
"grad_norm": 0.08093765377998352,
"learning_rate": 5.247867123947734e-05,
"loss": 0.7204088134765625,
"step": 35500
},
{
"epoch": 0.911949083905693,
"grad_norm": 0.07946806401014328,
"learning_rate": 4.893568776395227e-05,
"loss": 0.7201962280273437,
"step": 35750
},
{
"epoch": 0.9183263502267118,
"grad_norm": 0.07947609573602676,
"learning_rate": 4.53927042884272e-05,
"loss": 0.7209183349609375,
"step": 36000
},
{
"epoch": 0.9247036165477307,
"grad_norm": 0.07726657390594482,
"learning_rate": 4.184972081290213e-05,
"loss": 0.7201143188476562,
"step": 36250
},
{
"epoch": 0.9310808828687495,
"grad_norm": 0.08261343091726303,
"learning_rate": 3.8306737337377056e-05,
"loss": 0.7180040283203125,
"step": 36500
},
{
"epoch": 0.9374581491897683,
"grad_norm": 0.077142134308815,
"learning_rate": 3.476375386185199e-05,
"loss": 0.72004541015625,
"step": 36750
},
{
"epoch": 0.9438354155107872,
"grad_norm": 0.07826323062181473,
"learning_rate": 3.122077038632692e-05,
"loss": 0.718272705078125,
"step": 37000
},
{
"epoch": 0.950212681831806,
"grad_norm": 0.08283810317516327,
"learning_rate": 2.7677786910801846e-05,
"loss": 0.7193888549804688,
"step": 37250
},
{
"epoch": 0.9565899481528248,
"grad_norm": 0.07523984462022781,
"learning_rate": 2.4134803435276777e-05,
"loss": 0.7178954467773437,
"step": 37500
},
{
"epoch": 0.9629672144738436,
"grad_norm": 0.07884211093187332,
"learning_rate": 2.059181995975171e-05,
"loss": 0.7184935913085938,
"step": 37750
},
{
"epoch": 0.9693444807948625,
"grad_norm": 0.0814971998333931,
"learning_rate": 1.704883648422664e-05,
"loss": 0.7168883056640625,
"step": 38000
},
{
"epoch": 0.9757217471158813,
"grad_norm": 0.08023018389940262,
"learning_rate": 1.3505853008701568e-05,
"loss": 0.716118896484375,
"step": 38250
},
{
"epoch": 0.9820990134369001,
"grad_norm": 0.08309205621480942,
"learning_rate": 9.962869533176497e-06,
"loss": 0.7166826171875,
"step": 38500
},
{
"epoch": 0.988476279757919,
"grad_norm": 0.0776965394616127,
"learning_rate": 6.419886057651427e-06,
"loss": 0.7159559326171875,
"step": 38750
},
{
"epoch": 0.9948535460789378,
"grad_norm": 0.07634233683347702,
"learning_rate": 2.8769025821263568e-06,
"loss": 0.716369140625,
"step": 39000
},
{
"epoch": 1.0,
"step": 39202,
"total_flos": 7.001260646121603e+18,
"train_loss": 0.8267243484910438,
"train_runtime": 51831.3256,
"train_samples_per_second": 48.405,
"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": 7.001260646121603e+18,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}