test-cp90 / trainer_state.json
applet3's picture
Upload folder using huggingface_hub
6933ab1 verified
Raw History Blame Contribute Delete
17.4 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.5421686746987951,
"eval_steps": 500,
"global_step": 90,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006024096385542169,
"grad_norm": 11.142176628112793,
"learning_rate": 0.0,
"loss": 0.428902804851532,
"step": 1
},
{
"epoch": 0.012048192771084338,
"grad_norm": 11.034109115600586,
"learning_rate": 5.555555555555555e-07,
"loss": 0.42792809009552,
"step": 2
},
{
"epoch": 0.018072289156626505,
"grad_norm": 8.74612045288086,
"learning_rate": 1.111111111111111e-06,
"loss": 0.3810642957687378,
"step": 3
},
{
"epoch": 0.024096385542168676,
"grad_norm": 9.825540542602539,
"learning_rate": 1.6666666666666667e-06,
"loss": 0.4792688190937042,
"step": 4
},
{
"epoch": 0.030120481927710843,
"grad_norm": 9.189689636230469,
"learning_rate": 2.222222222222222e-06,
"loss": 0.40543031692504883,
"step": 5
},
{
"epoch": 0.03614457831325301,
"grad_norm": 7.393019199371338,
"learning_rate": 2.7777777777777783e-06,
"loss": 0.4046850800514221,
"step": 6
},
{
"epoch": 0.04216867469879518,
"grad_norm": 5.906346797943115,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.3256700038909912,
"step": 7
},
{
"epoch": 0.04819277108433735,
"grad_norm": 4.1642632484436035,
"learning_rate": 3.88888888888889e-06,
"loss": 0.2622697651386261,
"step": 8
},
{
"epoch": 0.05421686746987952,
"grad_norm": 4.239490032196045,
"learning_rate": 4.444444444444444e-06,
"loss": 0.31185054779052734,
"step": 9
},
{
"epoch": 0.060240963855421686,
"grad_norm": 2.696002721786499,
"learning_rate": 5e-06,
"loss": 0.26603323221206665,
"step": 10
},
{
"epoch": 0.06626506024096386,
"grad_norm": 2.6713006496429443,
"learning_rate": 4.999881749961716e-06,
"loss": 0.29827162623405457,
"step": 11
},
{
"epoch": 0.07228915662650602,
"grad_norm": 2.532520294189453,
"learning_rate": 4.999527011033321e-06,
"loss": 0.2619355022907257,
"step": 12
},
{
"epoch": 0.0783132530120482,
"grad_norm": 2.390526533126831,
"learning_rate": 4.9989358167731285e-06,
"loss": 0.2530556321144104,
"step": 13
},
{
"epoch": 0.08433734939759036,
"grad_norm": 2.513714551925659,
"learning_rate": 4.998108223108133e-06,
"loss": 0.2906743884086609,
"step": 14
},
{
"epoch": 0.09036144578313253,
"grad_norm": 2.0978147983551025,
"learning_rate": 4.9970443083287224e-06,
"loss": 0.23413600027561188,
"step": 15
},
{
"epoch": 0.0963855421686747,
"grad_norm": 2.065729856491089,
"learning_rate": 4.995744173081265e-06,
"loss": 0.24827918410301208,
"step": 16
},
{
"epoch": 0.10240963855421686,
"grad_norm": 1.884876012802124,
"learning_rate": 4.994207940358597e-06,
"loss": 0.26246076822280884,
"step": 17
},
{
"epoch": 0.10843373493975904,
"grad_norm": 2.5683226585388184,
"learning_rate": 4.99243575548838e-06,
"loss": 0.3206707835197449,
"step": 18
},
{
"epoch": 0.1144578313253012,
"grad_norm": 1.6468095779418945,
"learning_rate": 4.9904277861193574e-06,
"loss": 0.20793572068214417,
"step": 19
},
{
"epoch": 0.12048192771084337,
"grad_norm": 2.026458263397217,
"learning_rate": 4.988184222205492e-06,
"loss": 0.23015552759170532,
"step": 20
},
{
"epoch": 0.12650602409638553,
"grad_norm": 1.8717762231826782,
"learning_rate": 4.9857052759880006e-06,
"loss": 0.25441068410873413,
"step": 21
},
{
"epoch": 0.13253012048192772,
"grad_norm": 2.0290212631225586,
"learning_rate": 4.98299118197527e-06,
"loss": 0.22494634985923767,
"step": 22
},
{
"epoch": 0.13855421686746988,
"grad_norm": 2.214284658432007,
"learning_rate": 4.980042196920677e-06,
"loss": 0.26744288206100464,
"step": 23
},
{
"epoch": 0.14457831325301204,
"grad_norm": 2.1135001182556152,
"learning_rate": 4.976858599798299e-06,
"loss": 0.27789992094039917,
"step": 24
},
{
"epoch": 0.15060240963855423,
"grad_norm": 1.9528898000717163,
"learning_rate": 4.97344069177652e-06,
"loss": 0.2466088980436325,
"step": 25
},
{
"epoch": 0.1566265060240964,
"grad_norm": 2.101134777069092,
"learning_rate": 4.969788796189544e-06,
"loss": 0.2756834030151367,
"step": 26
},
{
"epoch": 0.16265060240963855,
"grad_norm": 2.343674421310425,
"learning_rate": 4.965903258506806e-06,
"loss": 0.28239595890045166,
"step": 27
},
{
"epoch": 0.1686746987951807,
"grad_norm": 2.253816843032837,
"learning_rate": 4.9617844463002895e-06,
"loss": 0.2833530306816101,
"step": 28
},
{
"epoch": 0.1746987951807229,
"grad_norm": 2.0298430919647217,
"learning_rate": 4.957432749209755e-06,
"loss": 0.25803640484809875,
"step": 29
},
{
"epoch": 0.18072289156626506,
"grad_norm": 1.9530489444732666,
"learning_rate": 4.952848578905881e-06,
"loss": 0.2215431183576584,
"step": 30
},
{
"epoch": 0.18674698795180722,
"grad_norm": 1.940727949142456,
"learning_rate": 4.948032369051317e-06,
"loss": 0.21956567466259003,
"step": 31
},
{
"epoch": 0.1927710843373494,
"grad_norm": 1.7422584295272827,
"learning_rate": 4.942984575259666e-06,
"loss": 0.20336072146892548,
"step": 32
},
{
"epoch": 0.19879518072289157,
"grad_norm": 1.7944018840789795,
"learning_rate": 4.937705675052373e-06,
"loss": 0.2340494990348816,
"step": 33
},
{
"epoch": 0.20481927710843373,
"grad_norm": 1.9949196577072144,
"learning_rate": 4.932196167813559e-06,
"loss": 0.30935412645339966,
"step": 34
},
{
"epoch": 0.21084337349397592,
"grad_norm": 1.9301334619522095,
"learning_rate": 4.926456574742778e-06,
"loss": 0.24893832206726074,
"step": 35
},
{
"epoch": 0.21686746987951808,
"grad_norm": 1.830209493637085,
"learning_rate": 4.920487438805711e-06,
"loss": 0.24357952177524567,
"step": 36
},
{
"epoch": 0.22289156626506024,
"grad_norm": 1.988722324371338,
"learning_rate": 4.9142893246828e-06,
"loss": 0.22917479276657104,
"step": 37
},
{
"epoch": 0.2289156626506024,
"grad_norm": 1.8187060356140137,
"learning_rate": 4.907862818715831e-06,
"loss": 0.27028727531433105,
"step": 38
},
{
"epoch": 0.23493975903614459,
"grad_norm": 2.018294095993042,
"learning_rate": 4.901208528852465e-06,
"loss": 0.2310202419757843,
"step": 39
},
{
"epoch": 0.24096385542168675,
"grad_norm": 2.0524044036865234,
"learning_rate": 4.894327084588727e-06,
"loss": 0.2731711268424988,
"step": 40
},
{
"epoch": 0.2469879518072289,
"grad_norm": 1.8044854402542114,
"learning_rate": 4.887219136909455e-06,
"loss": 0.21416833996772766,
"step": 41
},
{
"epoch": 0.25301204819277107,
"grad_norm": 1.7802573442459106,
"learning_rate": 4.879885358226718e-06,
"loss": 0.19698399305343628,
"step": 42
},
{
"epoch": 0.25903614457831325,
"grad_norm": 2.2974190711975098,
"learning_rate": 4.872326442316203e-06,
"loss": 0.27448707818984985,
"step": 43
},
{
"epoch": 0.26506024096385544,
"grad_norm": 2.2874867916107178,
"learning_rate": 4.864543104251587e-06,
"loss": 0.25334182381629944,
"step": 44
},
{
"epoch": 0.2710843373493976,
"grad_norm": 1.9537646770477295,
"learning_rate": 4.856536080336889e-06,
"loss": 0.23862236738204956,
"step": 45
},
{
"epoch": 0.27710843373493976,
"grad_norm": 2.0207152366638184,
"learning_rate": 4.848306128036817e-06,
"loss": 0.21768009662628174,
"step": 46
},
{
"epoch": 0.28313253012048195,
"grad_norm": 1.8584778308868408,
"learning_rate": 4.839854025905111e-06,
"loss": 0.22923576831817627,
"step": 47
},
{
"epoch": 0.2891566265060241,
"grad_norm": 1.950571060180664,
"learning_rate": 4.83118057351089e-06,
"loss": 0.24856369197368622,
"step": 48
},
{
"epoch": 0.29518072289156627,
"grad_norm": 1.9584742784500122,
"learning_rate": 4.822286591363018e-06,
"loss": 0.22143439948558807,
"step": 49
},
{
"epoch": 0.30120481927710846,
"grad_norm": 1.7273216247558594,
"learning_rate": 4.813172920832477e-06,
"loss": 0.21043998003005981,
"step": 50
},
{
"epoch": 0.3072289156626506,
"grad_norm": 1.895005702972412,
"learning_rate": 4.8038404240727785e-06,
"loss": 0.2245764136314392,
"step": 51
},
{
"epoch": 0.3132530120481928,
"grad_norm": 2.0731658935546875,
"learning_rate": 4.794289983938403e-06,
"loss": 0.22999663650989532,
"step": 52
},
{
"epoch": 0.3192771084337349,
"grad_norm": 1.8308407068252563,
"learning_rate": 4.784522503901279e-06,
"loss": 0.195350781083107,
"step": 53
},
{
"epoch": 0.3253012048192771,
"grad_norm": 1.843331217765808,
"learning_rate": 4.774538907965318e-06,
"loss": 0.2074728161096573,
"step": 54
},
{
"epoch": 0.3313253012048193,
"grad_norm": 1.7621232271194458,
"learning_rate": 4.764340140578999e-06,
"loss": 0.2011730819940567,
"step": 55
},
{
"epoch": 0.3373493975903614,
"grad_norm": 2.133352518081665,
"learning_rate": 4.753927166546032e-06,
"loss": 0.2272721230983734,
"step": 56
},
{
"epoch": 0.3433734939759036,
"grad_norm": 1.7030085325241089,
"learning_rate": 4.743300970934077e-06,
"loss": 0.21829667687416077,
"step": 57
},
{
"epoch": 0.3493975903614458,
"grad_norm": 1.9551599025726318,
"learning_rate": 4.732462558981566e-06,
"loss": 0.27412763237953186,
"step": 58
},
{
"epoch": 0.35542168674698793,
"grad_norm": 1.7485458850860596,
"learning_rate": 4.7214129560026015e-06,
"loss": 0.2115289866924286,
"step": 59
},
{
"epoch": 0.3614457831325301,
"grad_norm": 2.274949789047241,
"learning_rate": 4.710153207289963e-06,
"loss": 0.2607116997241974,
"step": 60
},
{
"epoch": 0.3674698795180723,
"grad_norm": 1.6567832231521606,
"learning_rate": 4.698684378016223e-06,
"loss": 0.2039959728717804,
"step": 61
},
{
"epoch": 0.37349397590361444,
"grad_norm": 2.048426866531372,
"learning_rate": 4.687007553132984e-06,
"loss": 0.2799915075302124,
"step": 62
},
{
"epoch": 0.3795180722891566,
"grad_norm": 1.7995165586471558,
"learning_rate": 4.675123837268235e-06,
"loss": 0.21062196791172028,
"step": 63
},
{
"epoch": 0.3855421686746988,
"grad_norm": 1.8946254253387451,
"learning_rate": 4.663034354621864e-06,
"loss": 0.20754952728748322,
"step": 64
},
{
"epoch": 0.39156626506024095,
"grad_norm": 1.7134093046188354,
"learning_rate": 4.650740248859297e-06,
"loss": 0.21463951468467712,
"step": 65
},
{
"epoch": 0.39759036144578314,
"grad_norm": 1.9219496250152588,
"learning_rate": 4.638242683003317e-06,
"loss": 0.26067647337913513,
"step": 66
},
{
"epoch": 0.4036144578313253,
"grad_norm": 1.7691570520401,
"learning_rate": 4.625542839324036e-06,
"loss": 0.23639249801635742,
"step": 67
},
{
"epoch": 0.40963855421686746,
"grad_norm": 1.9222103357315063,
"learning_rate": 4.6126419192270555e-06,
"loss": 0.2021777331829071,
"step": 68
},
{
"epoch": 0.41566265060240964,
"grad_norm": 1.7899214029312134,
"learning_rate": 4.599541143139811e-06,
"loss": 0.24930432438850403,
"step": 69
},
{
"epoch": 0.42168674698795183,
"grad_norm": 2.099332094192505,
"learning_rate": 4.586241750396123e-06,
"loss": 0.22903309762477875,
"step": 70
},
{
"epoch": 0.42771084337349397,
"grad_norm": 1.7237921953201294,
"learning_rate": 4.572744999118951e-06,
"loss": 0.21340343356132507,
"step": 71
},
{
"epoch": 0.43373493975903615,
"grad_norm": 2.0501701831817627,
"learning_rate": 4.55905216610138e-06,
"loss": 0.2957020699977875,
"step": 72
},
{
"epoch": 0.4397590361445783,
"grad_norm": 1.9058964252471924,
"learning_rate": 4.545164546685833e-06,
"loss": 0.1939191371202469,
"step": 73
},
{
"epoch": 0.4457831325301205,
"grad_norm": 1.6917678117752075,
"learning_rate": 4.531083454641531e-06,
"loss": 0.17142519354820251,
"step": 74
},
{
"epoch": 0.45180722891566266,
"grad_norm": 1.7717102766036987,
"learning_rate": 4.5168102220402134e-06,
"loss": 0.2252553552389145,
"step": 75
},
{
"epoch": 0.4578313253012048,
"grad_norm": 1.658211588859558,
"learning_rate": 4.502346199130122e-06,
"loss": 0.23183956742286682,
"step": 76
},
{
"epoch": 0.463855421686747,
"grad_norm": 1.8013274669647217,
"learning_rate": 4.487692754208266e-06,
"loss": 0.22619429230690002,
"step": 77
},
{
"epoch": 0.46987951807228917,
"grad_norm": 1.7795627117156982,
"learning_rate": 4.472851273490985e-06,
"loss": 0.2013862431049347,
"step": 78
},
{
"epoch": 0.4759036144578313,
"grad_norm": 1.939047932624817,
"learning_rate": 4.457823160982807e-06,
"loss": 0.24375547468662262,
"step": 79
},
{
"epoch": 0.4819277108433735,
"grad_norm": 2.0435917377471924,
"learning_rate": 4.442609838343639e-06,
"loss": 0.2810860276222229,
"step": 80
},
{
"epoch": 0.4879518072289157,
"grad_norm": 1.6849004030227661,
"learning_rate": 4.427212744754265e-06,
"loss": 0.20575395226478577,
"step": 81
},
{
"epoch": 0.4939759036144578,
"grad_norm": 1.7432836294174194,
"learning_rate": 4.411633336780213e-06,
"loss": 0.24015243351459503,
"step": 82
},
{
"epoch": 0.5,
"grad_norm": 2.005021810531616,
"learning_rate": 4.395873088233954e-06,
"loss": 0.2686704397201538,
"step": 83
},
{
"epoch": 0.5060240963855421,
"grad_norm": 1.9157148599624634,
"learning_rate": 4.379933490035482e-06,
"loss": 0.2602766752243042,
"step": 84
},
{
"epoch": 0.5120481927710844,
"grad_norm": 1.8026031255722046,
"learning_rate": 4.363816050071275e-06,
"loss": 0.2138879895210266,
"step": 85
},
{
"epoch": 0.5180722891566265,
"grad_norm": 1.8697198629379272,
"learning_rate": 4.3475222930516484e-06,
"loss": 0.24326106905937195,
"step": 86
},
{
"epoch": 0.5240963855421686,
"grad_norm": 1.840242624282837,
"learning_rate": 4.331053760366514e-06,
"loss": 0.23169377446174622,
"step": 87
},
{
"epoch": 0.5301204819277109,
"grad_norm": 1.9570574760437012,
"learning_rate": 4.314412009939568e-06,
"loss": 0.21268793940544128,
"step": 88
},
{
"epoch": 0.536144578313253,
"grad_norm": 6.490395545959473,
"learning_rate": 4.297598616080912e-06,
"loss": 0.23041915893554688,
"step": 89
},
{
"epoch": 0.5421686746987951,
"grad_norm": 1.701446294784546,
"learning_rate": 4.28061516933812e-06,
"loss": 0.19419308006763458,
"step": 90
}
],
"logging_steps": 1,
"max_steps": 332,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 30,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 5.1649109246450074e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}