mistral5pGrad / trainer_state.json
terry69's picture
Model save
176fa6b verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 406,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0024630541871921183,
"grad_norm": 0.13177792727947235,
"learning_rate": 4.8780487804878055e-06,
"loss": 0.7592,
"step": 1
},
{
"epoch": 0.012315270935960592,
"grad_norm": 0.1387966126203537,
"learning_rate": 2.4390243902439026e-05,
"loss": 0.7841,
"step": 5
},
{
"epoch": 0.024630541871921183,
"grad_norm": 0.13816028833389282,
"learning_rate": 4.878048780487805e-05,
"loss": 0.7896,
"step": 10
},
{
"epoch": 0.03694581280788178,
"grad_norm": 0.12815995514392853,
"learning_rate": 7.317073170731707e-05,
"loss": 0.7837,
"step": 15
},
{
"epoch": 0.04926108374384237,
"grad_norm": 0.11007481813430786,
"learning_rate": 9.75609756097561e-05,
"loss": 0.7807,
"step": 20
},
{
"epoch": 0.06157635467980296,
"grad_norm": 0.10207259654998779,
"learning_rate": 0.00012195121951219512,
"loss": 0.7567,
"step": 25
},
{
"epoch": 0.07389162561576355,
"grad_norm": 0.12121258676052094,
"learning_rate": 0.00014634146341463414,
"loss": 0.7527,
"step": 30
},
{
"epoch": 0.08620689655172414,
"grad_norm": 0.11584752053022385,
"learning_rate": 0.0001707317073170732,
"loss": 0.7514,
"step": 35
},
{
"epoch": 0.09852216748768473,
"grad_norm": 0.10999231040477753,
"learning_rate": 0.0001951219512195122,
"loss": 0.7499,
"step": 40
},
{
"epoch": 0.11083743842364532,
"grad_norm": 0.10557198524475098,
"learning_rate": 0.0001999407400739705,
"loss": 0.7379,
"step": 45
},
{
"epoch": 0.12315270935960591,
"grad_norm": 0.10717561095952988,
"learning_rate": 0.00019970011699250152,
"loss": 0.7747,
"step": 50
},
{
"epoch": 0.1354679802955665,
"grad_norm": 0.09578366577625275,
"learning_rate": 0.00019927487224577402,
"loss": 0.747,
"step": 55
},
{
"epoch": 0.1477832512315271,
"grad_norm": 0.09655096381902695,
"learning_rate": 0.0001986657932891657,
"loss": 0.7435,
"step": 60
},
{
"epoch": 0.16009852216748768,
"grad_norm": 0.09785150736570358,
"learning_rate": 0.00019787400799669154,
"loss": 0.7481,
"step": 65
},
{
"epoch": 0.1724137931034483,
"grad_norm": 0.10019391030073166,
"learning_rate": 0.00019690098257244064,
"loss": 0.7549,
"step": 70
},
{
"epoch": 0.18472906403940886,
"grad_norm": 0.10015880316495895,
"learning_rate": 0.00019574851883550395,
"loss": 0.7501,
"step": 75
},
{
"epoch": 0.19704433497536947,
"grad_norm": 0.10224644094705582,
"learning_rate": 0.00019441875088341997,
"loss": 0.7438,
"step": 80
},
{
"epoch": 0.20935960591133004,
"grad_norm": 0.09352747350931168,
"learning_rate": 0.00019291414114031743,
"loss": 0.7374,
"step": 85
},
{
"epoch": 0.22167487684729065,
"grad_norm": 0.09862257540225983,
"learning_rate": 0.00019123747579707275,
"loss": 0.7588,
"step": 90
},
{
"epoch": 0.23399014778325122,
"grad_norm": 0.09364981204271317,
"learning_rate": 0.0001893918596519257,
"loss": 0.7631,
"step": 95
},
{
"epoch": 0.24630541871921183,
"grad_norm": 0.10517935454845428,
"learning_rate": 0.00018738071036110808,
"loss": 0.7507,
"step": 100
},
{
"epoch": 0.25862068965517243,
"grad_norm": 0.10279802232980728,
"learning_rate": 0.00018520775211013093,
"loss": 0.7503,
"step": 105
},
{
"epoch": 0.270935960591133,
"grad_norm": 0.0953611433506012,
"learning_rate": 0.00018287700871745036,
"loss": 0.7644,
"step": 110
},
{
"epoch": 0.2832512315270936,
"grad_norm": 0.10334543138742447,
"learning_rate": 0.00018039279618328212,
"loss": 0.7601,
"step": 115
},
{
"epoch": 0.2955665024630542,
"grad_norm": 0.10075850784778595,
"learning_rate": 0.0001777597146973627,
"loss": 0.7391,
"step": 120
},
{
"epoch": 0.3078817733990148,
"grad_norm": 0.09204421937465668,
"learning_rate": 0.00017498264012045687,
"loss": 0.7442,
"step": 125
},
{
"epoch": 0.32019704433497537,
"grad_norm": 0.09441586583852768,
"learning_rate": 0.00017206671495538612,
"loss": 0.7407,
"step": 130
},
{
"epoch": 0.33251231527093594,
"grad_norm": 0.10386330634355545,
"learning_rate": 0.0001690173388242972,
"loss": 0.7371,
"step": 135
},
{
"epoch": 0.3448275862068966,
"grad_norm": 0.09605494141578674,
"learning_rate": 0.0001658401584698049,
"loss": 0.7467,
"step": 140
},
{
"epoch": 0.35714285714285715,
"grad_norm": 0.10094594210386276,
"learning_rate": 0.00016254105729852464,
"loss": 0.7414,
"step": 145
},
{
"epoch": 0.3694581280788177,
"grad_norm": 0.09067249298095703,
"learning_rate": 0.00015912614448635782,
"loss": 0.753,
"step": 150
},
{
"epoch": 0.3817733990147783,
"grad_norm": 0.10036266595125198,
"learning_rate": 0.00015560174366570446,
"loss": 0.7397,
"step": 155
},
{
"epoch": 0.39408866995073893,
"grad_norm": 0.10051223635673523,
"learning_rate": 0.0001519743812155516,
"loss": 0.7535,
"step": 160
},
{
"epoch": 0.4064039408866995,
"grad_norm": 0.09287708252668381,
"learning_rate": 0.00014825077417612186,
"loss": 0.754,
"step": 165
},
{
"epoch": 0.4187192118226601,
"grad_norm": 0.09688938409090042,
"learning_rate": 0.00014443781781046136,
"loss": 0.7584,
"step": 170
},
{
"epoch": 0.43103448275862066,
"grad_norm": 0.09463707357645035,
"learning_rate": 0.00014054257283599973,
"loss": 0.7192,
"step": 175
},
{
"epoch": 0.4433497536945813,
"grad_norm": 0.0989103689789772,
"learning_rate": 0.00013657225234972695,
"loss": 0.7346,
"step": 180
},
{
"epoch": 0.45566502463054187,
"grad_norm": 0.09697972238063812,
"learning_rate": 0.00013253420847119803,
"loss": 0.7356,
"step": 185
},
{
"epoch": 0.46798029556650245,
"grad_norm": 0.09980908036231995,
"learning_rate": 0.0001284359187281004,
"loss": 0.7403,
"step": 190
},
{
"epoch": 0.4802955665024631,
"grad_norm": 0.09531185775995255,
"learning_rate": 0.0001242849722095936,
"loss": 0.7283,
"step": 195
},
{
"epoch": 0.49261083743842365,
"grad_norm": 0.0942201018333435,
"learning_rate": 0.00012008905551306356,
"loss": 0.7513,
"step": 200
},
{
"epoch": 0.5049261083743842,
"grad_norm": 0.09356020390987396,
"learning_rate": 0.00011585593851031347,
"loss": 0.7375,
"step": 205
},
{
"epoch": 0.5172413793103449,
"grad_norm": 0.09078332781791687,
"learning_rate": 0.00011159345995955006,
"loss": 0.7341,
"step": 210
},
{
"epoch": 0.5295566502463054,
"grad_norm": 0.09174405783414841,
"learning_rate": 0.00010730951298980776,
"loss": 0.7338,
"step": 215
},
{
"epoch": 0.541871921182266,
"grad_norm": 0.0960531234741211,
"learning_rate": 0.00010301203048469083,
"loss": 0.7583,
"step": 220
},
{
"epoch": 0.5541871921182266,
"grad_norm": 0.10194765776395798,
"learning_rate": 9.870897039249911e-05,
"loss": 0.7339,
"step": 225
},
{
"epoch": 0.5665024630541872,
"grad_norm": 0.09580555558204651,
"learning_rate": 9.440830098993969e-05,
"loss": 0.7369,
"step": 230
},
{
"epoch": 0.5788177339901478,
"grad_norm": 0.09715011715888977,
"learning_rate": 9.011798612671286e-05,
"loss": 0.7232,
"step": 235
},
{
"epoch": 0.5911330049261084,
"grad_norm": 0.09561636298894882,
"learning_rate": 8.58459704782957e-05,
"loss": 0.7325,
"step": 240
},
{
"epoch": 0.603448275862069,
"grad_norm": 0.09226809442043304,
"learning_rate": 8.160016483423199e-05,
"loss": 0.7411,
"step": 245
},
{
"epoch": 0.6157635467980296,
"grad_norm": 0.09844076633453369,
"learning_rate": 7.738843144917119e-05,
"loss": 0.7347,
"step": 250
},
{
"epoch": 0.6280788177339901,
"grad_norm": 0.09305092692375183,
"learning_rate": 7.321856948378259e-05,
"loss": 0.7461,
"step": 255
},
{
"epoch": 0.6403940886699507,
"grad_norm": 0.09724168479442596,
"learning_rate": 6.909830056250527e-05,
"loss": 0.749,
"step": 260
},
{
"epoch": 0.6527093596059114,
"grad_norm": 0.09665533900260925,
"learning_rate": 6.503525447487715e-05,
"loss": 0.7365,
"step": 265
},
{
"epoch": 0.6650246305418719,
"grad_norm": 0.09020380675792694,
"learning_rate": 6.103695504692122e-05,
"loss": 0.7403,
"step": 270
},
{
"epoch": 0.6773399014778325,
"grad_norm": 0.09968876093626022,
"learning_rate": 5.7110806208751655e-05,
"loss": 0.7391,
"step": 275
},
{
"epoch": 0.6896551724137931,
"grad_norm": 0.09027230739593506,
"learning_rate": 5.326407828419979e-05,
"loss": 0.7473,
"step": 280
},
{
"epoch": 0.7019704433497537,
"grad_norm": 0.10408838093280792,
"learning_rate": 4.9503894527847964e-05,
"loss": 0.7405,
"step": 285
},
{
"epoch": 0.7142857142857143,
"grad_norm": 0.09234675765037537,
"learning_rate": 4.583721793440188e-05,
"loss": 0.7352,
"step": 290
},
{
"epoch": 0.7266009852216748,
"grad_norm": 0.09968586266040802,
"learning_rate": 4.227083834482728e-05,
"loss": 0.7563,
"step": 295
},
{
"epoch": 0.7389162561576355,
"grad_norm": 0.09535784274339676,
"learning_rate": 3.881135987312757e-05,
"loss": 0.7303,
"step": 300
},
{
"epoch": 0.7512315270935961,
"grad_norm": 0.09004243463277817,
"learning_rate": 3.546518867704499e-05,
"loss": 0.7462,
"step": 305
},
{
"epoch": 0.7635467980295566,
"grad_norm": 0.096128910779953,
"learning_rate": 3.223852109533112e-05,
"loss": 0.7392,
"step": 310
},
{
"epoch": 0.7758620689655172,
"grad_norm": 0.09637811779975891,
"learning_rate": 2.9137332173554043e-05,
"loss": 0.7318,
"step": 315
},
{
"epoch": 0.7881773399014779,
"grad_norm": 0.09532997012138367,
"learning_rate": 2.616736459968936e-05,
"loss": 0.7404,
"step": 320
},
{
"epoch": 0.8004926108374384,
"grad_norm": 0.0936516597867012,
"learning_rate": 2.33341180699841e-05,
"loss": 0.7556,
"step": 325
},
{
"epoch": 0.812807881773399,
"grad_norm": 0.0916033461689949,
"learning_rate": 2.0642839104785272e-05,
"loss": 0.7443,
"step": 330
},
{
"epoch": 0.8251231527093597,
"grad_norm": 0.09518487006425858,
"learning_rate": 1.8098511333192024e-05,
"loss": 0.7335,
"step": 335
},
{
"epoch": 0.8374384236453202,
"grad_norm": 0.09961594641208649,
"learning_rate": 1.570584626452173e-05,
"loss": 0.7602,
"step": 340
},
{
"epoch": 0.8497536945812808,
"grad_norm": 0.09134498238563538,
"learning_rate": 1.3469274563679402e-05,
"loss": 0.7456,
"step": 345
},
{
"epoch": 0.8620689655172413,
"grad_norm": 0.09200974553823471,
"learning_rate": 1.1392937846586215e-05,
"loss": 0.7409,
"step": 350
},
{
"epoch": 0.874384236453202,
"grad_norm": 0.09712038189172745,
"learning_rate": 9.48068101086026e-06,
"loss": 0.7192,
"step": 355
},
{
"epoch": 0.8866995073891626,
"grad_norm": 0.09649989008903503,
"learning_rate": 7.736045115951251e-06,
"loss": 0.7327,
"step": 360
},
{
"epoch": 0.8990147783251231,
"grad_norm": 0.09973736852407455,
"learning_rate": 6.16226082591359e-06,
"loss": 0.7295,
"step": 365
},
{
"epoch": 0.9113300492610837,
"grad_norm": 0.09414437413215637,
"learning_rate": 4.762242426960262e-06,
"loss": 0.7286,
"step": 370
},
{
"epoch": 0.9236453201970444,
"grad_norm": 0.08962981402873993,
"learning_rate": 3.5385824308756587e-06,
"loss": 0.7407,
"step": 375
},
{
"epoch": 0.9359605911330049,
"grad_norm": 0.09547747671604156,
"learning_rate": 2.493546774280531e-06,
"loss": 0.7322,
"step": 380
},
{
"epoch": 0.9482758620689655,
"grad_norm": 0.09376917779445648,
"learning_rate": 1.6290706226390285e-06,
"loss": 0.7347,
"step": 385
},
{
"epoch": 0.9605911330049262,
"grad_norm": 0.09975626319646835,
"learning_rate": 9.46754786777726e-07,
"loss": 0.7333,
"step": 390
},
{
"epoch": 0.9729064039408867,
"grad_norm": 0.09062483161687851,
"learning_rate": 4.4786275855247527e-07,
"loss": 0.7479,
"step": 395
},
{
"epoch": 0.9852216748768473,
"grad_norm": 0.09053165465593338,
"learning_rate": 1.333183711524133e-07,
"loss": 0.7375,
"step": 400
},
{
"epoch": 0.9975369458128078,
"grad_norm": 0.093317411839962,
"learning_rate": 3.7040883734462683e-09,
"loss": 0.7398,
"step": 405
},
{
"epoch": 1.0,
"eval_loss": 0.9064456224441528,
"eval_runtime": 996.7111,
"eval_samples_per_second": 1.159,
"eval_steps_per_second": 0.073,
"step": 406
},
{
"epoch": 1.0,
"step": 406,
"total_flos": 2.1410486778152878e+18,
"train_loss": 0.7449803688549643,
"train_runtime": 152626.4688,
"train_samples_per_second": 0.681,
"train_steps_per_second": 0.003
}
],
"logging_steps": 5,
"max_steps": 406,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 25,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.1410486778152878e+18,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}