nace2_level0 / trainer_state.json
mapama247's picture
upload nace2_level0 model
ff57cfa
{
"best_metric": 0.7953262197881779,
"best_model_checkpoint": "./output//roberta-large_nace_5__5e-6_0.01_0.06_07-12-22_21-12/checkpoint-98000",
"epoch": 0.08573862323490967,
"global_step": 100000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 2.8579512490675938e-08,
"loss": 0.6423,
"step": 2000
},
{
"epoch": 0.0,
"eval_accuracy": 0.00020930884194819268,
"eval_f1": 0.017752777264977686,
"eval_loss": 0.45688706636428833,
"eval_roc_auc": 0.4999505258927197,
"eval_runtime": 8888.0353,
"eval_samples_per_second": 33.327,
"step": 2000
},
{
"epoch": 0.0,
"learning_rate": 5.7159024981351875e-08,
"loss": 0.3833,
"step": 4000
},
{
"epoch": 0.0,
"eval_accuracy": 0.0,
"eval_f1": 0.0074795009434814345,
"eval_loss": 0.2739907205104828,
"eval_roc_auc": 0.4998445851936328,
"eval_runtime": 8865.9827,
"eval_samples_per_second": 33.41,
"step": 4000
},
{
"epoch": 0.01,
"learning_rate": 8.573853747202781e-08,
"loss": 0.2538,
"step": 6000
},
{
"epoch": 0.01,
"eval_accuracy": 0.00013166201348354057,
"eval_f1": 0.0004511461766688428,
"eval_loss": 0.2050675004720688,
"eval_roc_auc": 0.5000627296722496,
"eval_runtime": 8869.0926,
"eval_samples_per_second": 33.398,
"step": 6000
},
{
"epoch": 0.01,
"learning_rate": 1.1431804996270375e-07,
"loss": 0.1998,
"step": 8000
},
{
"epoch": 0.01,
"eval_accuracy": 3.3759490636805274e-06,
"eval_f1": 5.318554842278257e-06,
"eval_loss": 0.17527928948402405,
"eval_roc_auc": 0.4999982583562919,
"eval_runtime": 8867.7398,
"eval_samples_per_second": 33.403,
"step": 8000
},
{
"epoch": 0.01,
"learning_rate": 1.428975624533797e-07,
"loss": 0.1693,
"step": 10000
},
{
"epoch": 0.01,
"eval_accuracy": 0.18926583235712138,
"eval_f1": 0.310704534917664,
"eval_loss": 0.15055961906909943,
"eval_roc_auc": 0.5950524118809889,
"eval_runtime": 8868.9495,
"eval_samples_per_second": 33.399,
"step": 10000
},
{
"epoch": 0.01,
"learning_rate": 1.7147707494405562e-07,
"loss": 0.1457,
"step": 12000
},
{
"epoch": 0.01,
"eval_accuracy": 0.3331555333493128,
"eval_f1": 0.4681315118558145,
"eval_loss": 0.13044112920761108,
"eval_roc_auc": 0.6662536331303048,
"eval_runtime": 8874.2821,
"eval_samples_per_second": 33.379,
"step": 12000
},
{
"epoch": 0.01,
"learning_rate": 2.0005658743473155e-07,
"loss": 0.127,
"step": 14000
},
{
"epoch": 0.01,
"eval_accuracy": 0.41038374413006856,
"eval_f1": 0.5463731063419305,
"eval_loss": 0.11502532660961151,
"eval_roc_auc": 0.7051608184740583,
"eval_runtime": 8870.8129,
"eval_samples_per_second": 33.392,
"step": 14000
},
{
"epoch": 0.01,
"learning_rate": 2.286360999254075e-07,
"loss": 0.1132,
"step": 16000
},
{
"epoch": 0.01,
"eval_accuracy": 0.4763227812418766,
"eval_f1": 0.6070642556360302,
"eval_loss": 0.10388284176588058,
"eval_roc_auc": 0.7403745042255975,
"eval_runtime": 8872.294,
"eval_samples_per_second": 33.386,
"step": 16000
},
{
"epoch": 0.02,
"learning_rate": 2.5721561241608343e-07,
"loss": 0.1032,
"step": 18000
},
{
"epoch": 0.02,
"eval_accuracy": 0.5087960352854196,
"eval_f1": 0.6375540199236641,
"eval_loss": 0.09574878960847855,
"eval_roc_auc": 0.7589272821999232,
"eval_runtime": 8879.9353,
"eval_samples_per_second": 33.358,
"step": 18000
},
{
"epoch": 0.02,
"learning_rate": 2.857951249067594e-07,
"loss": 0.0959,
"step": 20000
},
{
"epoch": 0.02,
"eval_accuracy": 0.5288154132330452,
"eval_f1": 0.6573253473962186,
"eval_loss": 0.08954177796840668,
"eval_roc_auc": 0.7715952535178425,
"eval_runtime": 8893.417,
"eval_samples_per_second": 33.307,
"step": 20000
},
{
"epoch": 0.02,
"learning_rate": 3.1437463739743534e-07,
"loss": 0.0899,
"step": 22000
},
{
"epoch": 0.02,
"eval_accuracy": 0.5462656939432098,
"eval_f1": 0.6785063497828563,
"eval_loss": 0.08448488265275955,
"eval_roc_auc": 0.7872646974276402,
"eval_runtime": 8886.3548,
"eval_samples_per_second": 33.333,
"step": 22000
},
{
"epoch": 0.02,
"learning_rate": 3.4295414988811124e-07,
"loss": 0.085,
"step": 24000
},
{
"epoch": 0.02,
"eval_accuracy": 0.5589457586263938,
"eval_f1": 0.6890862787276392,
"eval_loss": 0.08046474307775497,
"eval_roc_auc": 0.793202863995178,
"eval_runtime": 8888.5138,
"eval_samples_per_second": 33.325,
"step": 24000
},
{
"epoch": 0.02,
"learning_rate": 3.715336623787872e-07,
"loss": 0.0807,
"step": 26000
},
{
"epoch": 0.02,
"eval_accuracy": 0.5779354721095968,
"eval_f1": 0.7085340529054537,
"eval_loss": 0.07694748789072037,
"eval_roc_auc": 0.8086207353956607,
"eval_runtime": 9108.0018,
"eval_samples_per_second": 32.522,
"step": 26000
},
{
"epoch": 0.02,
"learning_rate": 4.001131748694631e-07,
"loss": 0.0778,
"step": 28000
},
{
"epoch": 0.02,
"eval_accuracy": 0.5849236866714155,
"eval_f1": 0.7155799872243565,
"eval_loss": 0.07436001300811768,
"eval_roc_auc": 0.8137322286922238,
"eval_runtime": 8902.3443,
"eval_samples_per_second": 33.274,
"step": 28000
},
{
"epoch": 0.03,
"learning_rate": 4.28692687360139e-07,
"loss": 0.0753,
"step": 30000
},
{
"epoch": 0.03,
"eval_accuracy": 0.5928402872257463,
"eval_f1": 0.7241294650780317,
"eval_loss": 0.07221980392932892,
"eval_roc_auc": 0.8217651733984279,
"eval_runtime": 8918.1851,
"eval_samples_per_second": 33.214,
"step": 30000
},
{
"epoch": 0.03,
"learning_rate": 4.57272199850815e-07,
"loss": 0.0725,
"step": 32000
},
{
"epoch": 0.03,
"eval_accuracy": 0.5951764439778132,
"eval_f1": 0.7261386622802142,
"eval_loss": 0.07057042419910431,
"eval_roc_auc": 0.822130038839097,
"eval_runtime": 8995.4049,
"eval_samples_per_second": 32.929,
"step": 32000
},
{
"epoch": 0.03,
"learning_rate": 4.858517123414909e-07,
"loss": 0.0711,
"step": 34000
},
{
"epoch": 0.03,
"eval_accuracy": 0.6020971395583583,
"eval_f1": 0.7323235836003227,
"eval_loss": 0.06848303228616714,
"eval_roc_auc": 0.8254645532099296,
"eval_runtime": 9451.4989,
"eval_samples_per_second": 31.34,
"step": 34000
},
{
"epoch": 0.03,
"learning_rate": 5.144312248321669e-07,
"loss": 0.0696,
"step": 36000
},
{
"epoch": 0.03,
"eval_accuracy": 0.6094972199059461,
"eval_f1": 0.7393677046231971,
"eval_loss": 0.06683486700057983,
"eval_roc_auc": 0.8309554684971688,
"eval_runtime": 10018.5624,
"eval_samples_per_second": 29.566,
"step": 36000
},
{
"epoch": 0.03,
"learning_rate": 5.430107373228428e-07,
"loss": 0.0676,
"step": 38000
},
{
"epoch": 0.03,
"eval_accuracy": 0.6112223298774868,
"eval_f1": 0.7425602873268343,
"eval_loss": 0.06597442924976349,
"eval_roc_auc": 0.8349447666298531,
"eval_runtime": 11692.7897,
"eval_samples_per_second": 25.333,
"step": 38000
},
{
"epoch": 0.03,
"learning_rate": 5.715902498135188e-07,
"loss": 0.0663,
"step": 40000
},
{
"epoch": 0.03,
"eval_accuracy": 0.6157629813681371,
"eval_f1": 0.7460401136089835,
"eval_loss": 0.06471637636423111,
"eval_roc_auc": 0.8360501874414529,
"eval_runtime": 8963.8102,
"eval_samples_per_second": 33.045,
"step": 40000
},
{
"epoch": 0.04,
"learning_rate": 6.001697623041946e-07,
"loss": 0.0647,
"step": 42000
},
{
"epoch": 0.04,
"eval_accuracy": 0.6194090063569121,
"eval_f1": 0.7492256354462544,
"eval_loss": 0.06362345069646835,
"eval_roc_auc": 0.8375523158859597,
"eval_runtime": 8891.9369,
"eval_samples_per_second": 33.313,
"step": 42000
},
{
"epoch": 0.04,
"learning_rate": 6.287492747948707e-07,
"loss": 0.0642,
"step": 44000
},
{
"epoch": 0.04,
"eval_accuracy": 0.6211577479718986,
"eval_f1": 0.7510341371183263,
"eval_loss": 0.06275659799575806,
"eval_roc_auc": 0.8382498982808346,
"eval_runtime": 8889.7383,
"eval_samples_per_second": 33.321,
"step": 44000
},
{
"epoch": 0.04,
"learning_rate": 6.573287872855465e-07,
"loss": 0.0633,
"step": 46000
},
{
"epoch": 0.04,
"eval_accuracy": 0.6240340565741544,
"eval_f1": 0.7552442670824074,
"eval_loss": 0.061649855226278305,
"eval_roc_auc": 0.8417446851046798,
"eval_runtime": 9047.8013,
"eval_samples_per_second": 32.739,
"step": 46000
},
{
"epoch": 0.04,
"learning_rate": 6.859082997762225e-07,
"loss": 0.0623,
"step": 48000
},
{
"epoch": 0.04,
"eval_accuracy": 0.6283518954266019,
"eval_f1": 0.7614445531010631,
"eval_loss": 0.06095787137746811,
"eval_roc_auc": 0.8492713317355323,
"eval_runtime": 8895.4655,
"eval_samples_per_second": 33.299,
"step": 48000
},
{
"epoch": 0.04,
"learning_rate": 7.144878122668984e-07,
"loss": 0.0617,
"step": 50000
},
{
"epoch": 0.04,
"eval_accuracy": 0.6315151596992704,
"eval_f1": 0.7623575299697047,
"eval_loss": 0.059979602694511414,
"eval_roc_auc": 0.8468318365424115,
"eval_runtime": 8905.8909,
"eval_samples_per_second": 33.26,
"step": 50000
},
{
"epoch": 0.04,
"learning_rate": 7.430673247575744e-07,
"loss": 0.0603,
"step": 52000
},
{
"epoch": 0.04,
"eval_accuracy": 0.6323929064558274,
"eval_f1": 0.7646658582301569,
"eval_loss": 0.0595267117023468,
"eval_roc_auc": 0.8497986956155574,
"eval_runtime": 8916.2146,
"eval_samples_per_second": 33.222,
"step": 52000
},
{
"epoch": 0.05,
"learning_rate": 7.716468372482502e-07,
"loss": 0.0605,
"step": 54000
},
{
"epoch": 0.05,
"eval_accuracy": 0.6345197543659461,
"eval_f1": 0.7675632990904457,
"eval_loss": 0.05862889438867569,
"eval_roc_auc": 0.8518276428326589,
"eval_runtime": 8909.536,
"eval_samples_per_second": 33.247,
"step": 54000
},
{
"epoch": 0.05,
"learning_rate": 8.002263497389262e-07,
"loss": 0.0594,
"step": 56000
},
{
"epoch": 0.05,
"eval_accuracy": 0.6367647604932937,
"eval_f1": 0.7709907906594025,
"eval_loss": 0.05837056040763855,
"eval_roc_auc": 0.8562941632344473,
"eval_runtime": 8903.6442,
"eval_samples_per_second": 33.269,
"step": 56000
},
{
"epoch": 0.05,
"learning_rate": 8.288058622296021e-07,
"loss": 0.0592,
"step": 58000
},
{
"epoch": 0.05,
"eval_accuracy": 0.6382434261831857,
"eval_f1": 0.7715028842192007,
"eval_loss": 0.05767415836453438,
"eval_roc_auc": 0.8549143755897282,
"eval_runtime": 8915.2586,
"eval_samples_per_second": 33.225,
"step": 58000
},
{
"epoch": 0.05,
"learning_rate": 8.57385374720278e-07,
"loss": 0.0585,
"step": 60000
},
{
"epoch": 0.05,
"eval_accuracy": 0.6385371337517259,
"eval_f1": 0.7683973772822023,
"eval_loss": 0.057559411972761154,
"eval_roc_auc": 0.8502475677045068,
"eval_runtime": 8911.1959,
"eval_samples_per_second": 33.241,
"step": 60000
},
{
"epoch": 0.05,
"learning_rate": 8.85964887210954e-07,
"loss": 0.0578,
"step": 62000
},
{
"epoch": 0.05,
"eval_accuracy": 0.643915020610169,
"eval_f1": 0.7735556679913914,
"eval_loss": 0.056609489023685455,
"eval_roc_auc": 0.8534382569874794,
"eval_runtime": 8924.5002,
"eval_samples_per_second": 33.191,
"step": 62000
},
{
"epoch": 0.05,
"learning_rate": 9.1454439970163e-07,
"loss": 0.0572,
"step": 64000
},
{
"epoch": 0.05,
"eval_accuracy": 0.6444484205622305,
"eval_f1": 0.7776911375576832,
"eval_loss": 0.05652829632163048,
"eval_roc_auc": 0.8603552529596437,
"eval_runtime": 8931.0301,
"eval_samples_per_second": 33.167,
"step": 64000
},
{
"epoch": 0.06,
"learning_rate": 9.43123912192306e-07,
"loss": 0.0571,
"step": 66000
},
{
"epoch": 0.06,
"eval_accuracy": 0.6454882128738442,
"eval_f1": 0.7742083504686046,
"eval_loss": 0.05627002194523811,
"eval_roc_auc": 0.8534170877970823,
"eval_runtime": 8935.7869,
"eval_samples_per_second": 33.149,
"step": 66000
},
{
"epoch": 0.06,
"learning_rate": 9.717034246829818e-07,
"loss": 0.0563,
"step": 68000
},
{
"epoch": 0.06,
"eval_accuracy": 0.6462410495150449,
"eval_f1": 0.7776261820301417,
"eval_loss": 0.05584974214434624,
"eval_roc_auc": 0.8583744396413601,
"eval_runtime": 8945.9116,
"eval_samples_per_second": 33.112,
"step": 68000
},
{
"epoch": 0.06,
"learning_rate": 1.0002829371736579e-06,
"loss": 0.0562,
"step": 70000
},
{
"epoch": 0.06,
"eval_accuracy": 0.6459709735899505,
"eval_f1": 0.7794595088938681,
"eval_loss": 0.05508234724402428,
"eval_roc_auc": 0.8591985509814772,
"eval_runtime": 8944.2225,
"eval_samples_per_second": 33.118,
"step": 70000
},
{
"epoch": 0.06,
"learning_rate": 1.0288624496643337e-06,
"loss": 0.0554,
"step": 72000
},
{
"epoch": 0.06,
"eval_accuracy": 0.6490059517981993,
"eval_f1": 0.7775233968578572,
"eval_loss": 0.05504510924220085,
"eval_roc_auc": 0.8542740144934026,
"eval_runtime": 8953.1995,
"eval_samples_per_second": 33.085,
"step": 72000
},
{
"epoch": 0.06,
"learning_rate": 1.0574419621550096e-06,
"loss": 0.0552,
"step": 74000
},
{
"epoch": 0.06,
"eval_accuracy": 0.6509471225098156,
"eval_f1": 0.7816832192367912,
"eval_loss": 0.054713618010282516,
"eval_roc_auc": 0.86037433954253,
"eval_runtime": 8945.5727,
"eval_samples_per_second": 33.113,
"step": 74000
},
{
"epoch": 0.07,
"learning_rate": 1.0860214746456856e-06,
"loss": 0.0547,
"step": 76000
},
{
"epoch": 0.07,
"eval_accuracy": 0.6510315212364076,
"eval_f1": 0.783673992804474,
"eval_loss": 0.05407635122537613,
"eval_roc_auc": 0.8622274781842953,
"eval_runtime": 8946.7264,
"eval_samples_per_second": 33.109,
"step": 76000
},
{
"epoch": 0.07,
"learning_rate": 1.1146009871363615e-06,
"loss": 0.0547,
"step": 78000
},
{
"epoch": 0.07,
"eval_accuracy": 0.6545526361098264,
"eval_f1": 0.7855580273587336,
"eval_loss": 0.053795717656612396,
"eval_roc_auc": 0.863520124962409,
"eval_runtime": 8947.2041,
"eval_samples_per_second": 33.107,
"step": 78000
},
{
"epoch": 0.07,
"learning_rate": 1.1431804996270375e-06,
"loss": 0.0539,
"step": 80000
},
{
"epoch": 0.07,
"eval_accuracy": 0.6558962638371713,
"eval_f1": 0.7874432731515895,
"eval_loss": 0.053422119468450546,
"eval_roc_auc": 0.8649282174633808,
"eval_runtime": 8946.6771,
"eval_samples_per_second": 33.109,
"step": 80000
},
{
"epoch": 0.07,
"learning_rate": 1.1717600121177134e-06,
"loss": 0.0541,
"step": 82000
},
{
"epoch": 0.07,
"eval_accuracy": 0.654451357637916,
"eval_f1": 0.7878992158367831,
"eval_loss": 0.05306807905435562,
"eval_roc_auc": 0.8663645266466831,
"eval_runtime": 8946.194,
"eval_samples_per_second": 33.111,
"step": 82000
},
{
"epoch": 0.07,
"learning_rate": 1.2003395246083892e-06,
"loss": 0.0534,
"step": 84000
},
{
"epoch": 0.07,
"eval_accuracy": 0.6543129437263051,
"eval_f1": 0.7862359590552527,
"eval_loss": 0.05337332561612129,
"eval_roc_auc": 0.8637659543017016,
"eval_runtime": 8946.5504,
"eval_samples_per_second": 33.109,
"step": 84000
},
{
"epoch": 0.07,
"learning_rate": 1.2289190370990653e-06,
"loss": 0.0533,
"step": 86000
},
{
"epoch": 0.07,
"eval_accuracy": 0.6579623446641437,
"eval_f1": 0.789641153558776,
"eval_loss": 0.05264722555875778,
"eval_roc_auc": 0.8661964026287946,
"eval_runtime": 8946.6953,
"eval_samples_per_second": 33.109,
"step": 86000
},
{
"epoch": 0.08,
"learning_rate": 1.2574985495897414e-06,
"loss": 0.0536,
"step": 88000
},
{
"epoch": 0.08,
"eval_accuracy": 0.6578779459375517,
"eval_f1": 0.7904054501316083,
"eval_loss": 0.05252569913864136,
"eval_roc_auc": 0.8665749344562018,
"eval_runtime": 8946.467,
"eval_samples_per_second": 33.109,
"step": 88000
},
{
"epoch": 0.08,
"learning_rate": 1.2860780620804172e-06,
"loss": 0.0529,
"step": 90000
},
{
"epoch": 0.08,
"eval_accuracy": 0.6560279258506547,
"eval_f1": 0.7903261630505988,
"eval_loss": 0.052733857184648514,
"eval_roc_auc": 0.8701088706242622,
"eval_runtime": 8948.9618,
"eval_samples_per_second": 33.1,
"step": 90000
},
{
"epoch": 0.08,
"learning_rate": 1.314657574571093e-06,
"loss": 0.0525,
"step": 92000
},
{
"epoch": 0.08,
"eval_accuracy": 0.6622430480768906,
"eval_f1": 0.7920309543792765,
"eval_loss": 0.0518718883395195,
"eval_roc_auc": 0.8668831436253378,
"eval_runtime": 8946.4377,
"eval_samples_per_second": 33.11,
"step": 92000
},
{
"epoch": 0.08,
"learning_rate": 1.343237087061769e-06,
"loss": 0.0522,
"step": 94000
},
{
"epoch": 0.08,
"eval_accuracy": 0.6575808624199478,
"eval_f1": 0.7916439543928127,
"eval_loss": 0.052168361842632294,
"eval_roc_auc": 0.8691981685442582,
"eval_runtime": 8948.4665,
"eval_samples_per_second": 33.102,
"step": 94000
},
{
"epoch": 0.08,
"learning_rate": 1.371816599552445e-06,
"loss": 0.0518,
"step": 96000
},
{
"epoch": 0.08,
"eval_accuracy": 0.6617164000229565,
"eval_f1": 0.7952230824994735,
"eval_loss": 0.051213569939136505,
"eval_roc_auc": 0.8707353068301484,
"eval_runtime": 8946.6417,
"eval_samples_per_second": 33.109,
"step": 96000
},
{
"epoch": 0.08,
"learning_rate": 1.4003961120431208e-06,
"loss": 0.0515,
"step": 98000
},
{
"epoch": 0.08,
"eval_accuracy": 0.6638094884424384,
"eval_f1": 0.7953262197881779,
"eval_loss": 0.05120311677455902,
"eval_roc_auc": 0.8704717586188649,
"eval_runtime": 8948.3566,
"eval_samples_per_second": 33.103,
"step": 98000
},
{
"epoch": 0.09,
"learning_rate": 1.4289756245337969e-06,
"loss": 0.0515,
"step": 100000
},
{
"epoch": 0.09,
"eval_accuracy": 0.6594410103540358,
"eval_f1": 0.7952893225951836,
"eval_loss": 0.05145873501896858,
"eval_roc_auc": 0.8740886532218961,
"eval_runtime": 8946.2483,
"eval_samples_per_second": 33.11,
"step": 100000
}
],
"max_steps": 5831675,
"num_train_epochs": 5,
"total_flos": 6.9871808151552e+16,
"trial_name": null,
"trial_params": null
}