mistral-7b-ift / trainer_state.json
NicholasCorrado's picture
Model save
d62f00a verified
raw
history blame
17.2 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 461,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0021691973969631237,
"grad_norm": 11.479755487865775,
"learning_rate": 1.0638297872340425e-08,
"loss": 1.204,
"step": 1
},
{
"epoch": 0.010845986984815618,
"grad_norm": 11.719854924785,
"learning_rate": 5.3191489361702123e-08,
"loss": 1.2705,
"step": 5
},
{
"epoch": 0.021691973969631236,
"grad_norm": 12.5013105505897,
"learning_rate": 1.0638297872340425e-07,
"loss": 1.2569,
"step": 10
},
{
"epoch": 0.03253796095444685,
"grad_norm": 10.664645875250606,
"learning_rate": 1.5957446808510638e-07,
"loss": 1.2504,
"step": 15
},
{
"epoch": 0.04338394793926247,
"grad_norm": 9.933034533041308,
"learning_rate": 2.127659574468085e-07,
"loss": 1.2132,
"step": 20
},
{
"epoch": 0.05422993492407809,
"grad_norm": 5.664415659701916,
"learning_rate": 2.659574468085106e-07,
"loss": 1.1718,
"step": 25
},
{
"epoch": 0.0650759219088937,
"grad_norm": 4.799955074967526,
"learning_rate": 3.1914893617021275e-07,
"loss": 1.1492,
"step": 30
},
{
"epoch": 0.07592190889370933,
"grad_norm": 3.5056771066313868,
"learning_rate": 3.7234042553191484e-07,
"loss": 1.1193,
"step": 35
},
{
"epoch": 0.08676789587852494,
"grad_norm": 2.758357067802773,
"learning_rate": 4.25531914893617e-07,
"loss": 1.0995,
"step": 40
},
{
"epoch": 0.09761388286334056,
"grad_norm": 2.512596535973654,
"learning_rate": 4.787234042553192e-07,
"loss": 1.0789,
"step": 45
},
{
"epoch": 0.10845986984815618,
"grad_norm": 2.4219547091489244,
"learning_rate": 4.999352212103373e-07,
"loss": 1.0741,
"step": 50
},
{
"epoch": 0.1193058568329718,
"grad_norm": 2.359757061351952,
"learning_rate": 4.995394723941067e-07,
"loss": 1.0585,
"step": 55
},
{
"epoch": 0.1301518438177874,
"grad_norm": 2.4328040984593247,
"learning_rate": 4.98784531935359e-07,
"loss": 1.0568,
"step": 60
},
{
"epoch": 0.14099783080260303,
"grad_norm": 2.3052389483909543,
"learning_rate": 4.976714865090826e-07,
"loss": 1.0638,
"step": 65
},
{
"epoch": 0.15184381778741865,
"grad_norm": 2.198699363060523,
"learning_rate": 4.96201938253052e-07,
"loss": 1.0349,
"step": 70
},
{
"epoch": 0.16268980477223427,
"grad_norm": 2.138037840029951,
"learning_rate": 4.943780024616802e-07,
"loss": 1.034,
"step": 75
},
{
"epoch": 0.1735357917570499,
"grad_norm": 2.1989759782977076,
"learning_rate": 4.922023045412265e-07,
"loss": 1.0295,
"step": 80
},
{
"epoch": 0.1843817787418655,
"grad_norm": 2.2941860079446235,
"learning_rate": 4.896779762307389e-07,
"loss": 1.0262,
"step": 85
},
{
"epoch": 0.19522776572668113,
"grad_norm": 2.1481967771932773,
"learning_rate": 4.868086510941716e-07,
"loss": 1.0154,
"step": 90
},
{
"epoch": 0.20607375271149675,
"grad_norm": 2.160671871794234,
"learning_rate": 4.835984592901677e-07,
"loss": 1.0468,
"step": 95
},
{
"epoch": 0.21691973969631237,
"grad_norm": 2.1550419124695783,
"learning_rate": 4.800520216270341e-07,
"loss": 1.0336,
"step": 100
},
{
"epoch": 0.227765726681128,
"grad_norm": 2.2253199590666157,
"learning_rate": 4.7617444291146555e-07,
"loss": 1.0224,
"step": 105
},
{
"epoch": 0.2386117136659436,
"grad_norm": 2.183801883640947,
"learning_rate": 4.7197130460059377e-07,
"loss": 1.0139,
"step": 110
},
{
"epoch": 0.24945770065075923,
"grad_norm": 2.177485956300299,
"learning_rate": 4.6744865676793666e-07,
"loss": 1.0053,
"step": 115
},
{
"epoch": 0.2603036876355748,
"grad_norm": 2.1256298478006856,
"learning_rate": 4.6261300939481274e-07,
"loss": 0.9914,
"step": 120
},
{
"epoch": 0.27114967462039047,
"grad_norm": 2.1075218318516886,
"learning_rate": 4.574713229997563e-07,
"loss": 1.0006,
"step": 125
},
{
"epoch": 0.28199566160520606,
"grad_norm": 2.1304778495096457,
"learning_rate": 4.520309986194201e-07,
"loss": 1.0338,
"step": 130
},
{
"epoch": 0.2928416485900217,
"grad_norm": 2.1650488920006046,
"learning_rate": 4.462998671553897e-07,
"loss": 1.0017,
"step": 135
},
{
"epoch": 0.3036876355748373,
"grad_norm": 2.214789820606799,
"learning_rate": 4.4028617810224115e-07,
"loss": 1.0101,
"step": 140
},
{
"epoch": 0.31453362255965295,
"grad_norm": 2.157917041478876,
"learning_rate": 4.3399858767306927e-07,
"loss": 0.989,
"step": 145
},
{
"epoch": 0.32537960954446854,
"grad_norm": 2.092011368725107,
"learning_rate": 4.2744614633957723e-07,
"loss": 1.0115,
"step": 150
},
{
"epoch": 0.3362255965292842,
"grad_norm": 2.219786897391391,
"learning_rate": 4.206382858046635e-07,
"loss": 0.9853,
"step": 155
},
{
"epoch": 0.3470715835140998,
"grad_norm": 2.0979071528204756,
"learning_rate": 4.135848054262578e-07,
"loss": 1.0163,
"step": 160
},
{
"epoch": 0.3579175704989154,
"grad_norm": 2.1698498942112288,
"learning_rate": 4.062958581119472e-07,
"loss": 0.9865,
"step": 165
},
{
"epoch": 0.368763557483731,
"grad_norm": 2.132812386449239,
"learning_rate": 3.9878193570469743e-07,
"loss": 0.9871,
"step": 170
},
{
"epoch": 0.3796095444685466,
"grad_norm": 2.337071820270912,
"learning_rate": 3.91053853880703e-07,
"loss": 0.984,
"step": 175
},
{
"epoch": 0.39045553145336226,
"grad_norm": 2.1129666341045263,
"learning_rate": 3.831227365811074e-07,
"loss": 0.9943,
"step": 180
},
{
"epoch": 0.40130151843817785,
"grad_norm": 2.121223339662348,
"learning_rate": 3.75e-07,
"loss": 0.9918,
"step": 185
},
{
"epoch": 0.4121475054229935,
"grad_norm": 2.285212865590998,
"learning_rate": 3.6669733615173965e-07,
"loss": 0.9843,
"step": 190
},
{
"epoch": 0.4229934924078091,
"grad_norm": 2.123810421477716,
"learning_rate": 3.5822669604125684e-07,
"loss": 0.9781,
"step": 195
},
{
"epoch": 0.43383947939262474,
"grad_norm": 2.1627048249065997,
"learning_rate": 3.4960027246156036e-07,
"loss": 0.9781,
"step": 200
},
{
"epoch": 0.44468546637744033,
"grad_norm": 2.137516375815561,
"learning_rate": 3.408304824432103e-07,
"loss": 0.9867,
"step": 205
},
{
"epoch": 0.455531453362256,
"grad_norm": 2.058882247220635,
"learning_rate": 3.319299493810187e-07,
"loss": 0.9776,
"step": 210
},
{
"epoch": 0.46637744034707157,
"grad_norm": 2.1433762495027864,
"learning_rate": 3.229114848637062e-07,
"loss": 0.9879,
"step": 215
},
{
"epoch": 0.4772234273318872,
"grad_norm": 2.1877299182133734,
"learning_rate": 3.13788070232669e-07,
"loss": 0.9812,
"step": 220
},
{
"epoch": 0.4880694143167028,
"grad_norm": 2.0613909390774525,
"learning_rate": 3.0457283789640036e-07,
"loss": 1.0005,
"step": 225
},
{
"epoch": 0.49891540130151846,
"grad_norm": 2.137287960677187,
"learning_rate": 2.9527905242746395e-07,
"loss": 0.9994,
"step": 230
},
{
"epoch": 0.5097613882863341,
"grad_norm": 2.062311669373047,
"learning_rate": 2.85920091469227e-07,
"loss": 0.9806,
"step": 235
},
{
"epoch": 0.5206073752711496,
"grad_norm": 2.1009325069210054,
"learning_rate": 2.765094264798387e-07,
"loss": 0.9905,
"step": 240
},
{
"epoch": 0.5314533622559653,
"grad_norm": 2.252996820494139,
"learning_rate": 2.6706060334116775e-07,
"loss": 0.9833,
"step": 245
},
{
"epoch": 0.5422993492407809,
"grad_norm": 2.063111058570344,
"learning_rate": 2.575872228606156e-07,
"loss": 0.9466,
"step": 250
},
{
"epoch": 0.5531453362255966,
"grad_norm": 2.1028294210091794,
"learning_rate": 2.4810292119386674e-07,
"loss": 0.9556,
"step": 255
},
{
"epoch": 0.5639913232104121,
"grad_norm": 2.121237281068362,
"learning_rate": 2.3862135021675915e-07,
"loss": 0.9717,
"step": 260
},
{
"epoch": 0.5748373101952278,
"grad_norm": 2.2205226946736905,
"learning_rate": 2.2915615787452664e-07,
"loss": 0.9638,
"step": 265
},
{
"epoch": 0.5856832971800434,
"grad_norm": 2.0940361008010333,
"learning_rate": 2.1972096853669903e-07,
"loss": 0.9671,
"step": 270
},
{
"epoch": 0.596529284164859,
"grad_norm": 2.0586787374748576,
"learning_rate": 2.1032936338593717e-07,
"loss": 0.9773,
"step": 275
},
{
"epoch": 0.6073752711496746,
"grad_norm": 2.072790057342988,
"learning_rate": 2.0099486086903294e-07,
"loss": 0.9684,
"step": 280
},
{
"epoch": 0.6182212581344902,
"grad_norm": 2.1778824831003147,
"learning_rate": 1.9173089723821087e-07,
"loss": 0.9667,
"step": 285
},
{
"epoch": 0.6290672451193059,
"grad_norm": 2.115203742538656,
"learning_rate": 1.825508072107439e-07,
"loss": 0.9919,
"step": 290
},
{
"epoch": 0.6399132321041214,
"grad_norm": 2.9734190506035536,
"learning_rate": 1.7346780477471897e-07,
"loss": 0.9741,
"step": 295
},
{
"epoch": 0.6507592190889371,
"grad_norm": 2.109405285224148,
"learning_rate": 1.6449496416858282e-07,
"loss": 0.9815,
"step": 300
},
{
"epoch": 0.6616052060737527,
"grad_norm": 2.0612586919012355,
"learning_rate": 1.5564520106184643e-07,
"loss": 0.9637,
"step": 305
},
{
"epoch": 0.6724511930585684,
"grad_norm": 2.126282772097656,
"learning_rate": 1.4693125396403562e-07,
"loss": 0.9576,
"step": 310
},
{
"epoch": 0.6832971800433839,
"grad_norm": 2.151977422363772,
"learning_rate": 1.3836566588865e-07,
"loss": 0.9871,
"step": 315
},
{
"epoch": 0.6941431670281996,
"grad_norm": 2.10699457923422,
"learning_rate": 1.2996076629852112e-07,
"loss": 0.9775,
"step": 320
},
{
"epoch": 0.7049891540130152,
"grad_norm": 2.2306202203282366,
"learning_rate": 1.2172865335856064e-07,
"loss": 0.9869,
"step": 325
},
{
"epoch": 0.7158351409978309,
"grad_norm": 2.1309989319293696,
"learning_rate": 1.1368117652144185e-07,
"loss": 0.9871,
"step": 330
},
{
"epoch": 0.7266811279826464,
"grad_norm": 2.1871677586329175,
"learning_rate": 1.0582991947128323e-07,
"loss": 0.983,
"step": 335
},
{
"epoch": 0.737527114967462,
"grad_norm": 2.089837928751529,
"learning_rate": 9.818618344988258e-08,
"loss": 0.9817,
"step": 340
},
{
"epoch": 0.7483731019522777,
"grad_norm": 2.0529724920050905,
"learning_rate": 9.076097098950541e-08,
"loss": 0.9581,
"step": 345
},
{
"epoch": 0.7592190889370932,
"grad_norm": 2.077217246150324,
"learning_rate": 8.356497007563986e-08,
"loss": 0.9845,
"step": 350
},
{
"epoch": 0.7700650759219089,
"grad_norm": 2.1176827367409876,
"learning_rate": 7.660853876251683e-08,
"loss": 0.9625,
"step": 355
},
{
"epoch": 0.7809110629067245,
"grad_norm": 2.1409566690457433,
"learning_rate": 6.990169026353867e-08,
"loss": 0.9999,
"step": 360
},
{
"epoch": 0.7917570498915402,
"grad_norm": 2.0254717410055703,
"learning_rate": 6.345407853807863e-08,
"loss": 0.9642,
"step": 365
},
{
"epoch": 0.8026030368763557,
"grad_norm": 2.140899328755212,
"learning_rate": 5.727498439539602e-08,
"loss": 0.9717,
"step": 370
},
{
"epoch": 0.8134490238611713,
"grad_norm": 2.0666287819389404,
"learning_rate": 5.13733021356714e-08,
"loss": 0.9697,
"step": 375
},
{
"epoch": 0.824295010845987,
"grad_norm": 2.0741290971837127,
"learning_rate": 4.57575267473895e-08,
"loss": 0.9641,
"step": 380
},
{
"epoch": 0.8351409978308026,
"grad_norm": 2.108990519804528,
"learning_rate": 4.043574167949892e-08,
"loss": 0.9902,
"step": 385
},
{
"epoch": 0.8459869848156182,
"grad_norm": 2.094620559269544,
"learning_rate": 3.541560720594869e-08,
"loss": 0.9926,
"step": 390
},
{
"epoch": 0.8568329718004338,
"grad_norm": 2.0594257251140586,
"learning_rate": 3.0704349399351435e-08,
"loss": 0.9614,
"step": 395
},
{
"epoch": 0.8676789587852495,
"grad_norm": 2.1291510379356753,
"learning_rate": 2.6308749729643058e-08,
"loss": 0.9651,
"step": 400
},
{
"epoch": 0.8785249457700651,
"grad_norm": 2.1035620640567734,
"learning_rate": 2.2235135302712092e-08,
"loss": 0.9952,
"step": 405
},
{
"epoch": 0.8893709327548807,
"grad_norm": 2.0424482935650614,
"learning_rate": 1.8489369753048682e-08,
"loss": 0.9631,
"step": 410
},
{
"epoch": 0.9002169197396963,
"grad_norm": 2.0939056533766847,
"learning_rate": 1.507684480352292e-08,
"loss": 0.9747,
"step": 415
},
{
"epoch": 0.911062906724512,
"grad_norm": 4.429046210547233,
"learning_rate": 1.2002472504440807e-08,
"loss": 0.9843,
"step": 420
},
{
"epoch": 0.9219088937093276,
"grad_norm": 2.1047314550346297,
"learning_rate": 9.270678163050217e-09,
"loss": 0.9738,
"step": 425
},
{
"epoch": 0.9327548806941431,
"grad_norm": 2.0998363541587888,
"learning_rate": 6.885393973673298e-09,
"loss": 0.9509,
"step": 430
},
{
"epoch": 0.9436008676789588,
"grad_norm": 2.138035512894866,
"learning_rate": 4.850053357634693e-09,
"loss": 0.9799,
"step": 435
},
{
"epoch": 0.9544468546637744,
"grad_norm": 2.053994471901985,
"learning_rate": 3.1675860211325954e-09,
"loss": 0.9762,
"step": 440
},
{
"epoch": 0.96529284164859,
"grad_norm": 2.054984790757136,
"learning_rate": 1.840413738166402e-09,
"loss": 0.9499,
"step": 445
},
{
"epoch": 0.9761388286334056,
"grad_norm": 2.0710147404144097,
"learning_rate": 8.704468645914787e-10,
"loss": 0.9692,
"step": 450
},
{
"epoch": 0.9869848156182213,
"grad_norm": 2.067619347023652,
"learning_rate": 2.5908158831811077e-10,
"loss": 0.9994,
"step": 455
},
{
"epoch": 0.9978308026030369,
"grad_norm": 2.095822455094207,
"learning_rate": 7.197919613455284e-12,
"loss": 0.9524,
"step": 460
},
{
"epoch": 1.0,
"eval_loss": 1.0288481712341309,
"eval_runtime": 3.4782,
"eval_samples_per_second": 74.465,
"eval_steps_per_second": 1.438,
"step": 461
},
{
"epoch": 1.0,
"step": 461,
"total_flos": 192943352709120.0,
"train_loss": 1.0081440616326323,
"train_runtime": 5423.7345,
"train_samples_per_second": 21.752,
"train_steps_per_second": 0.085
}
],
"logging_steps": 5,
"max_steps": 461,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 192943352709120.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}