test-cp210 / trainer_state.json
applet3's picture
Upload folder using huggingface_hub
2a0f5ff verified
Raw History Blame Contribute Delete
39.6 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.2650602409638554,
"eval_steps": 500,
"global_step": 210,
"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
},
{
"epoch": 0.5481927710843374,
"grad_norm": 1.8265910148620605,
"learning_rate": 4.2634632763457725e-06,
"loss": 0.21390798687934875,
"step": 91
},
{
"epoch": 0.5542168674698795,
"grad_norm": 1.8848633766174316,
"learning_rate": 4.246144559673472e-06,
"loss": 0.234243705868721,
"step": 92
},
{
"epoch": 0.5602409638554217,
"grad_norm": 1.739385724067688,
"learning_rate": 4.228660657672347e-06,
"loss": 0.197153240442276,
"step": 93
},
{
"epoch": 0.5662650602409639,
"grad_norm": 1.7461135387420654,
"learning_rate": 4.211013224320062e-06,
"loss": 0.19059306383132935,
"step": 94
},
{
"epoch": 0.572289156626506,
"grad_norm": 1.9738885164260864,
"learning_rate": 4.1932039290643534e-06,
"loss": 0.21974927186965942,
"step": 95
},
{
"epoch": 0.5783132530120482,
"grad_norm": 1.7048437595367432,
"learning_rate": 4.175234456665096e-06,
"loss": 0.1835278868675232,
"step": 96
},
{
"epoch": 0.5843373493975904,
"grad_norm": 1.8517098426818848,
"learning_rate": 4.15710650703493e-06,
"loss": 0.2452908456325531,
"step": 97
},
{
"epoch": 0.5903614457831325,
"grad_norm": 1.8890876770019531,
"learning_rate": 4.138821795078447e-06,
"loss": 0.2501678466796875,
"step": 98
},
{
"epoch": 0.5963855421686747,
"grad_norm": 1.751680850982666,
"learning_rate": 4.120382050529957e-06,
"loss": 0.21254503726959229,
"step": 99
},
{
"epoch": 0.6024096385542169,
"grad_norm": 1.9866863489151,
"learning_rate": 4.101789017789858e-06,
"loss": 0.26266080141067505,
"step": 100
},
{
"epoch": 0.608433734939759,
"grad_norm": 2.1752734184265137,
"learning_rate": 4.083044455759618e-06,
"loss": 0.2721603214740753,
"step": 101
},
{
"epoch": 0.6144578313253012,
"grad_norm": 1.7891639471054077,
"learning_rate": 4.064150137675378e-06,
"loss": 0.21918655931949615,
"step": 102
},
{
"epoch": 0.6204819277108434,
"grad_norm": 1.7796125411987305,
"learning_rate": 4.045107850940209e-06,
"loss": 0.28184622526168823,
"step": 103
},
{
"epoch": 0.6265060240963856,
"grad_norm": 1.5150421857833862,
"learning_rate": 4.025919396955019e-06,
"loss": 0.1914132535457611,
"step": 104
},
{
"epoch": 0.6325301204819277,
"grad_norm": 1.752693772315979,
"learning_rate": 4.006586590948141e-06,
"loss": 0.23128381371498108,
"step": 105
},
{
"epoch": 0.6385542168674698,
"grad_norm": 1.8561952114105225,
"learning_rate": 3.987111261803618e-06,
"loss": 0.22575634717941284,
"step": 106
},
{
"epoch": 0.6445783132530121,
"grad_norm": 1.6240315437316895,
"learning_rate": 3.967495251888181e-06,
"loss": 0.20704583823680878,
"step": 107
},
{
"epoch": 0.6506024096385542,
"grad_norm": 1.8144125938415527,
"learning_rate": 3.947740416876971e-06,
"loss": 0.20163677632808685,
"step": 108
},
{
"epoch": 0.6566265060240963,
"grad_norm": 1.8921997547149658,
"learning_rate": 3.9278486255779835e-06,
"loss": 0.21554049849510193,
"step": 109
},
{
"epoch": 0.6626506024096386,
"grad_norm": 1.804047703742981,
"learning_rate": 3.907821759755286e-06,
"loss": 0.23601463437080383,
"step": 110
},
{
"epoch": 0.6686746987951807,
"grad_norm": 1.6722337007522583,
"learning_rate": 3.8876617139509965e-06,
"loss": 0.25775057077407837,
"step": 111
},
{
"epoch": 0.6746987951807228,
"grad_norm": 2.067570686340332,
"learning_rate": 3.8673703953060685e-06,
"loss": 0.23134759068489075,
"step": 112
},
{
"epoch": 0.6807228915662651,
"grad_norm": 1.8785966634750366,
"learning_rate": 3.846949723379863e-06,
"loss": 0.23591741919517517,
"step": 113
},
{
"epoch": 0.6867469879518072,
"grad_norm": 2.127025604248047,
"learning_rate": 3.826401629968575e-06,
"loss": 0.23402580618858337,
"step": 114
},
{
"epoch": 0.6927710843373494,
"grad_norm": 1.544934868812561,
"learning_rate": 3.805728058922467e-06,
"loss": 0.1696513295173645,
"step": 115
},
{
"epoch": 0.6987951807228916,
"grad_norm": 1.9435012340545654,
"learning_rate": 3.7849309659619943e-06,
"loss": 0.224832683801651,
"step": 116
},
{
"epoch": 0.7048192771084337,
"grad_norm": 1.7729871273040771,
"learning_rate": 3.7640123184927876e-06,
"loss": 0.21986815333366394,
"step": 117
},
{
"epoch": 0.7108433734939759,
"grad_norm": 1.636151671409607,
"learning_rate": 3.742974095419538e-06,
"loss": 0.19982966780662537,
"step": 118
},
{
"epoch": 0.7168674698795181,
"grad_norm": 1.822206735610962,
"learning_rate": 3.7218182869587923e-06,
"loss": 0.25640130043029785,
"step": 119
},
{
"epoch": 0.7228915662650602,
"grad_norm": 1.7592462301254272,
"learning_rate": 3.70054689445068e-06,
"loss": 0.24146384000778198,
"step": 120
},
{
"epoch": 0.7289156626506024,
"grad_norm": 2.0406506061553955,
"learning_rate": 3.6791619301695824e-06,
"loss": 0.24204370379447937,
"step": 121
},
{
"epoch": 0.7349397590361446,
"grad_norm": 1.798247218132019,
"learning_rate": 3.6576654171337765e-06,
"loss": 0.19507844746112823,
"step": 122
},
{
"epoch": 0.7409638554216867,
"grad_norm": 1.9995570182800293,
"learning_rate": 3.6360593889140543e-06,
"loss": 0.20422318577766418,
"step": 123
},
{
"epoch": 0.7469879518072289,
"grad_norm": 2.01058030128479,
"learning_rate": 3.6143458894413463e-06,
"loss": 0.24596446752548218,
"step": 124
},
{
"epoch": 0.7530120481927711,
"grad_norm": 1.6475582122802734,
"learning_rate": 3.592526972813367e-06,
"loss": 0.20919951796531677,
"step": 125
},
{
"epoch": 0.7590361445783133,
"grad_norm": 1.663503646850586,
"learning_rate": 3.5706047031002993e-06,
"loss": 0.21953055262565613,
"step": 126
},
{
"epoch": 0.7650602409638554,
"grad_norm": 1.7317630052566528,
"learning_rate": 3.5485811541495274e-06,
"loss": 0.2076019048690796,
"step": 127
},
{
"epoch": 0.7710843373493976,
"grad_norm": 1.492385983467102,
"learning_rate": 3.526458409389459e-06,
"loss": 0.17310795187950134,
"step": 128
},
{
"epoch": 0.7771084337349398,
"grad_norm": 1.6732460260391235,
"learning_rate": 3.5042385616324243e-06,
"loss": 0.1909010410308838,
"step": 129
},
{
"epoch": 0.7831325301204819,
"grad_norm": 1.8162691593170166,
"learning_rate": 3.4819237128767012e-06,
"loss": 0.22556881606578827,
"step": 130
},
{
"epoch": 0.7891566265060241,
"grad_norm": 1.8092666864395142,
"learning_rate": 3.4595159741076674e-06,
"loss": 0.22908125817775726,
"step": 131
},
{
"epoch": 0.7951807228915663,
"grad_norm": 1.759559154510498,
"learning_rate": 3.4370174650980953e-06,
"loss": 0.22184421122074127,
"step": 132
},
{
"epoch": 0.8012048192771084,
"grad_norm": 1.816255807876587,
"learning_rate": 3.4144303142076268e-06,
"loss": 0.20111408829689026,
"step": 133
},
{
"epoch": 0.8072289156626506,
"grad_norm": 1.6726490259170532,
"learning_rate": 3.3917566581814277e-06,
"loss": 0.20235317945480347,
"step": 134
},
{
"epoch": 0.8132530120481928,
"grad_norm": 1.8278712034225464,
"learning_rate": 3.368998641948052e-06,
"loss": 0.25044453144073486,
"step": 135
},
{
"epoch": 0.8192771084337349,
"grad_norm": 1.8829212188720703,
"learning_rate": 3.346158418416533e-06,
"loss": 0.24094229936599731,
"step": 136
},
{
"epoch": 0.8253012048192772,
"grad_norm": 1.904211401939392,
"learning_rate": 3.323238148272717e-06,
"loss": 0.22016692161560059,
"step": 137
},
{
"epoch": 0.8313253012048193,
"grad_norm": 1.8119757175445557,
"learning_rate": 3.3002399997748604e-06,
"loss": 0.2539440989494324,
"step": 138
},
{
"epoch": 0.8373493975903614,
"grad_norm": 1.6660523414611816,
"learning_rate": 3.277166148548516e-06,
"loss": 0.21818749606609344,
"step": 139
},
{
"epoch": 0.8433734939759037,
"grad_norm": 1.9667959213256836,
"learning_rate": 3.2540187773807162e-06,
"loss": 0.21603018045425415,
"step": 140
},
{
"epoch": 0.8493975903614458,
"grad_norm": 1.7938474416732788,
"learning_rate": 3.230800076013483e-06,
"loss": 0.20440860092639923,
"step": 141
},
{
"epoch": 0.8554216867469879,
"grad_norm": 1.7303977012634277,
"learning_rate": 3.207512240936676e-06,
"loss": 0.2181878238916397,
"step": 142
},
{
"epoch": 0.8614457831325302,
"grad_norm": 1.7015107870101929,
"learning_rate": 3.184157475180208e-06,
"loss": 0.23690566420555115,
"step": 143
},
{
"epoch": 0.8674698795180723,
"grad_norm": 1.7471257448196411,
"learning_rate": 3.1607379881056333e-06,
"loss": 0.22546926140785217,
"step": 144
},
{
"epoch": 0.8734939759036144,
"grad_norm": 1.7830501794815063,
"learning_rate": 3.137255995197147e-06,
"loss": 0.19030848145484924,
"step": 145
},
{
"epoch": 0.8795180722891566,
"grad_norm": 1.5704936981201172,
"learning_rate": 3.1137137178519983e-06,
"loss": 0.21496745944023132,
"step": 146
},
{
"epoch": 0.8855421686746988,
"grad_norm": 1.9630683660507202,
"learning_rate": 3.090113383170344e-06,
"loss": 0.2366226762533188,
"step": 147
},
{
"epoch": 0.891566265060241,
"grad_norm": 1.7654658555984497,
"learning_rate": 3.0664572237445684e-06,
"loss": 0.26164698600769043,
"step": 148
},
{
"epoch": 0.8975903614457831,
"grad_norm": 1.6233384609222412,
"learning_rate": 3.042747477448078e-06,
"loss": 0.17893335223197937,
"step": 149
},
{
"epoch": 0.9036144578313253,
"grad_norm": 1.801803708076477,
"learning_rate": 3.018986387223597e-06,
"loss": 0.22494494915008545,
"step": 150
},
{
"epoch": 0.9096385542168675,
"grad_norm": 1.7251133918762207,
"learning_rate": 0.0,
"loss": 0.21339407563209534,
"step": 151
},
{
"epoch": 0.9156626506024096,
"grad_norm": 1.9012843370437622,
"learning_rate": 5.555555555555555e-07,
"loss": 0.24071244895458221,
"step": 152
},
{
"epoch": 0.9216867469879518,
"grad_norm": 1.6273796558380127,
"learning_rate": 1.111111111111111e-06,
"loss": 0.19139668345451355,
"step": 153
},
{
"epoch": 0.927710843373494,
"grad_norm": 1.7698148488998413,
"learning_rate": 1.6666666666666667e-06,
"loss": 0.22368842363357544,
"step": 154
},
{
"epoch": 0.9337349397590361,
"grad_norm": 1.7884771823883057,
"learning_rate": 2.222222222222222e-06,
"loss": 0.16862092912197113,
"step": 155
},
{
"epoch": 0.9397590361445783,
"grad_norm": 1.6387245655059814,
"learning_rate": 2.7777777777777783e-06,
"loss": 0.24389740824699402,
"step": 156
},
{
"epoch": 0.9457831325301205,
"grad_norm": 1.9914135932922363,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.22333872318267822,
"step": 157
},
{
"epoch": 0.9518072289156626,
"grad_norm": 1.6643177270889282,
"learning_rate": 3.88888888888889e-06,
"loss": 0.20086194574832916,
"step": 158
},
{
"epoch": 0.9578313253012049,
"grad_norm": 1.6900769472122192,
"learning_rate": 4.444444444444444e-06,
"loss": 0.20764178037643433,
"step": 159
},
{
"epoch": 0.963855421686747,
"grad_norm": 1.817714810371399,
"learning_rate": 5e-06,
"loss": 0.24200844764709473,
"step": 160
},
{
"epoch": 0.9698795180722891,
"grad_norm": 1.7363492250442505,
"learning_rate": 4.999881749961716e-06,
"loss": 0.207527294754982,
"step": 161
},
{
"epoch": 0.9759036144578314,
"grad_norm": 1.7865567207336426,
"learning_rate": 4.999527011033321e-06,
"loss": 0.21617457270622253,
"step": 162
},
{
"epoch": 0.9819277108433735,
"grad_norm": 1.9564707279205322,
"learning_rate": 4.9989358167731285e-06,
"loss": 0.2597506642341614,
"step": 163
},
{
"epoch": 0.9879518072289156,
"grad_norm": 1.88765549659729,
"learning_rate": 4.998108223108133e-06,
"loss": 0.2530207931995392,
"step": 164
},
{
"epoch": 0.9939759036144579,
"grad_norm": 1.703033685684204,
"learning_rate": 4.9970443083287224e-06,
"loss": 0.18831050395965576,
"step": 165
},
{
"epoch": 1.0,
"grad_norm": 1.7099109888076782,
"learning_rate": 4.995744173081265e-06,
"loss": 0.21984407305717468,
"step": 166
},
{
"epoch": 1.0060240963855422,
"grad_norm": 1.407286524772644,
"learning_rate": 4.994207940358597e-06,
"loss": 0.09358582645654678,
"step": 167
},
{
"epoch": 1.0120481927710843,
"grad_norm": 1.4928256273269653,
"learning_rate": 4.99243575548838e-06,
"loss": 0.10626605153083801,
"step": 168
},
{
"epoch": 1.0180722891566265,
"grad_norm": 1.4833334684371948,
"learning_rate": 4.9904277861193574e-06,
"loss": 0.12156569212675095,
"step": 169
},
{
"epoch": 1.0240963855421688,
"grad_norm": 1.4212719202041626,
"learning_rate": 4.988184222205492e-06,
"loss": 0.10794398933649063,
"step": 170
},
{
"epoch": 1.0301204819277108,
"grad_norm": 1.5113886594772339,
"learning_rate": 4.9857052759880006e-06,
"loss": 0.12738291919231415,
"step": 171
},
{
"epoch": 1.036144578313253,
"grad_norm": 1.4146442413330078,
"learning_rate": 4.98299118197527e-06,
"loss": 0.09423454105854034,
"step": 172
},
{
"epoch": 1.0421686746987953,
"grad_norm": 1.4581801891326904,
"learning_rate": 4.980042196920677e-06,
"loss": 0.0916857123374939,
"step": 173
},
{
"epoch": 1.0481927710843373,
"grad_norm": 1.7278649806976318,
"learning_rate": 4.976858599798299e-06,
"loss": 0.1625499427318573,
"step": 174
},
{
"epoch": 1.0542168674698795,
"grad_norm": 1.7814223766326904,
"learning_rate": 4.97344069177652e-06,
"loss": 0.09106528013944626,
"step": 175
},
{
"epoch": 1.0602409638554218,
"grad_norm": 1.7452236413955688,
"learning_rate": 4.969788796189544e-06,
"loss": 0.09855592250823975,
"step": 176
},
{
"epoch": 1.0662650602409638,
"grad_norm": 1.877186894416809,
"learning_rate": 4.965903258506806e-06,
"loss": 0.11333008855581284,
"step": 177
},
{
"epoch": 1.072289156626506,
"grad_norm": 1.9699362516403198,
"learning_rate": 4.9617844463002895e-06,
"loss": 0.10632774233818054,
"step": 178
},
{
"epoch": 1.0783132530120483,
"grad_norm": 1.8559452295303345,
"learning_rate": 4.957432749209755e-06,
"loss": 0.09790239483118057,
"step": 179
},
{
"epoch": 1.0843373493975903,
"grad_norm": 2.022092342376709,
"learning_rate": 4.952848578905881e-06,
"loss": 0.13124418258666992,
"step": 180
},
{
"epoch": 1.0903614457831325,
"grad_norm": 1.9110645055770874,
"learning_rate": 4.948032369051317e-06,
"loss": 0.11170981824398041,
"step": 181
},
{
"epoch": 1.0963855421686748,
"grad_norm": 1.678166151046753,
"learning_rate": 4.942984575259666e-06,
"loss": 0.10503833740949631,
"step": 182
},
{
"epoch": 1.1024096385542168,
"grad_norm": 1.5391852855682373,
"learning_rate": 4.937705675052373e-06,
"loss": 0.10899101197719574,
"step": 183
},
{
"epoch": 1.108433734939759,
"grad_norm": 1.842432975769043,
"learning_rate": 4.932196167813559e-06,
"loss": 0.1137915700674057,
"step": 184
},
{
"epoch": 1.1144578313253013,
"grad_norm": 1.6912317276000977,
"learning_rate": 4.926456574742778e-06,
"loss": 0.11944588273763657,
"step": 185
},
{
"epoch": 1.1204819277108433,
"grad_norm": 1.6168071031570435,
"learning_rate": 4.920487438805711e-06,
"loss": 0.10693936049938202,
"step": 186
},
{
"epoch": 1.1265060240963856,
"grad_norm": 1.6483373641967773,
"learning_rate": 4.9142893246828e-06,
"loss": 0.12357063591480255,
"step": 187
},
{
"epoch": 1.1325301204819278,
"grad_norm": 1.4628320932388306,
"learning_rate": 4.907862818715831e-06,
"loss": 0.12019909918308258,
"step": 188
},
{
"epoch": 1.1385542168674698,
"grad_norm": 1.4515324831008911,
"learning_rate": 4.901208528852465e-06,
"loss": 0.10171081125736237,
"step": 189
},
{
"epoch": 1.144578313253012,
"grad_norm": 1.558293104171753,
"learning_rate": 4.894327084588727e-06,
"loss": 0.09913239628076553,
"step": 190
},
{
"epoch": 1.1506024096385543,
"grad_norm": 1.6200357675552368,
"learning_rate": 4.887219136909455e-06,
"loss": 0.12148313224315643,
"step": 191
},
{
"epoch": 1.1566265060240963,
"grad_norm": 1.3701298236846924,
"learning_rate": 4.879885358226718e-06,
"loss": 0.10428446531295776,
"step": 192
},
{
"epoch": 1.1626506024096386,
"grad_norm": 1.5008060932159424,
"learning_rate": 4.872326442316203e-06,
"loss": 0.08828811347484589,
"step": 193
},
{
"epoch": 1.1686746987951806,
"grad_norm": 1.4441299438476562,
"learning_rate": 4.864543104251587e-06,
"loss": 0.11046949028968811,
"step": 194
},
{
"epoch": 1.1746987951807228,
"grad_norm": 1.5599843263626099,
"learning_rate": 4.856536080336889e-06,
"loss": 0.1164817363023758,
"step": 195
},
{
"epoch": 1.180722891566265,
"grad_norm": 1.5171695947647095,
"learning_rate": 4.848306128036817e-06,
"loss": 0.0939384251832962,
"step": 196
},
{
"epoch": 1.1867469879518073,
"grad_norm": 1.586988091468811,
"learning_rate": 4.839854025905111e-06,
"loss": 0.12363364547491074,
"step": 197
},
{
"epoch": 1.1927710843373494,
"grad_norm": 1.7755016088485718,
"learning_rate": 4.83118057351089e-06,
"loss": 0.14537136256694794,
"step": 198
},
{
"epoch": 1.1987951807228916,
"grad_norm": 2.2900898456573486,
"learning_rate": 4.822286591363018e-06,
"loss": 0.12790751457214355,
"step": 199
},
{
"epoch": 1.2048192771084336,
"grad_norm": 1.7060179710388184,
"learning_rate": 4.813172920832477e-06,
"loss": 0.10709622502326965,
"step": 200
},
{
"epoch": 1.2108433734939759,
"grad_norm": 1.8671693801879883,
"learning_rate": 4.8038404240727785e-06,
"loss": 0.08105172216892242,
"step": 201
},
{
"epoch": 1.216867469879518,
"grad_norm": 1.736363172531128,
"learning_rate": 4.794289983938403e-06,
"loss": 0.1407506763935089,
"step": 202
},
{
"epoch": 1.2228915662650603,
"grad_norm": 1.4807357788085938,
"learning_rate": 4.784522503901279e-06,
"loss": 0.10849492251873016,
"step": 203
},
{
"epoch": 1.2289156626506024,
"grad_norm": 1.6700947284698486,
"learning_rate": 4.774538907965318e-06,
"loss": 0.13219457864761353,
"step": 204
},
{
"epoch": 1.2349397590361446,
"grad_norm": 1.740336298942566,
"learning_rate": 4.764340140578999e-06,
"loss": 0.1016375720500946,
"step": 205
},
{
"epoch": 1.2409638554216866,
"grad_norm": 1.567693829536438,
"learning_rate": 4.753927166546032e-06,
"loss": 0.08787193894386292,
"step": 206
},
{
"epoch": 1.2469879518072289,
"grad_norm": 1.3981975317001343,
"learning_rate": 4.743300970934077e-06,
"loss": 0.074513278901577,
"step": 207
},
{
"epoch": 1.2530120481927711,
"grad_norm": 1.5041234493255615,
"learning_rate": 4.732462558981566e-06,
"loss": 0.10591389983892441,
"step": 208
},
{
"epoch": 1.2590361445783134,
"grad_norm": 1.7367664575576782,
"learning_rate": 4.7214129560026015e-06,
"loss": 0.12603846192359924,
"step": 209
},
{
"epoch": 1.2650602409638554,
"grad_norm": 1.6155883073806763,
"learning_rate": 4.710153207289963e-06,
"loss": 0.0956132709980011,
"step": 210
}
],
"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": 1.2021715907211428e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}