2275 lines
55 KiB
JSON
2275 lines
55 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 9.872057318321392,
|
|
"eval_steps": 500,
|
|
"global_step": 1600,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.030706243602865915,
|
|
"grad_norm": 4.6859283447265625,
|
|
"learning_rate": 1.6460905349794242e-07,
|
|
"loss": 0.6431,
|
|
"step": 5
|
|
},
|
|
{
|
|
"epoch": 0.06141248720573183,
|
|
"grad_norm": 4.602360725402832,
|
|
"learning_rate": 3.7037037037037036e-07,
|
|
"loss": 0.6447,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.09211873080859775,
|
|
"grad_norm": 4.142496585845947,
|
|
"learning_rate": 5.761316872427984e-07,
|
|
"loss": 0.6395,
|
|
"step": 15
|
|
},
|
|
{
|
|
"epoch": 0.12282497441146366,
|
|
"grad_norm": 3.5161383152008057,
|
|
"learning_rate": 7.818930041152265e-07,
|
|
"loss": 0.6226,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.1535312180143296,
|
|
"grad_norm": 1.9937578439712524,
|
|
"learning_rate": 9.876543209876544e-07,
|
|
"loss": 0.5901,
|
|
"step": 25
|
|
},
|
|
{
|
|
"epoch": 0.1842374616171955,
|
|
"grad_norm": 1.0944325923919678,
|
|
"learning_rate": 1.1934156378600823e-06,
|
|
"loss": 0.575,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.21494370522006143,
|
|
"grad_norm": 0.8474246859550476,
|
|
"learning_rate": 1.3991769547325104e-06,
|
|
"loss": 0.5552,
|
|
"step": 35
|
|
},
|
|
{
|
|
"epoch": 0.24564994882292732,
|
|
"grad_norm": 0.869394838809967,
|
|
"learning_rate": 1.6049382716049383e-06,
|
|
"loss": 0.5416,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.27635619242579323,
|
|
"grad_norm": 0.6358171105384827,
|
|
"learning_rate": 1.8106995884773665e-06,
|
|
"loss": 0.534,
|
|
"step": 45
|
|
},
|
|
{
|
|
"epoch": 0.3070624360286592,
|
|
"grad_norm": 0.5201377868652344,
|
|
"learning_rate": 2.0164609053497946e-06,
|
|
"loss": 0.5236,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.33776867963152507,
|
|
"grad_norm": 0.44874125719070435,
|
|
"learning_rate": 2.222222222222222e-06,
|
|
"loss": 0.5176,
|
|
"step": 55
|
|
},
|
|
{
|
|
"epoch": 0.368474923234391,
|
|
"grad_norm": 0.33835747838020325,
|
|
"learning_rate": 2.4279835390946504e-06,
|
|
"loss": 0.5123,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.3991811668372569,
|
|
"grad_norm": 0.32708311080932617,
|
|
"learning_rate": 2.6337448559670788e-06,
|
|
"loss": 0.5092,
|
|
"step": 65
|
|
},
|
|
{
|
|
"epoch": 0.42988741044012285,
|
|
"grad_norm": 0.2938404083251953,
|
|
"learning_rate": 2.8395061728395062e-06,
|
|
"loss": 0.5046,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 0.46059365404298874,
|
|
"grad_norm": 0.2861787974834442,
|
|
"learning_rate": 3.0452674897119346e-06,
|
|
"loss": 0.4985,
|
|
"step": 75
|
|
},
|
|
{
|
|
"epoch": 0.49129989764585463,
|
|
"grad_norm": 0.26230937242507935,
|
|
"learning_rate": 3.2510288065843625e-06,
|
|
"loss": 0.4994,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 0.5220061412487206,
|
|
"grad_norm": 0.2605642080307007,
|
|
"learning_rate": 3.4567901234567904e-06,
|
|
"loss": 0.499,
|
|
"step": 85
|
|
},
|
|
{
|
|
"epoch": 0.5527123848515865,
|
|
"grad_norm": 0.2520701289176941,
|
|
"learning_rate": 3.6625514403292183e-06,
|
|
"loss": 0.4969,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 0.5834186284544524,
|
|
"grad_norm": 0.2760030925273895,
|
|
"learning_rate": 3.868312757201647e-06,
|
|
"loss": 0.4972,
|
|
"step": 95
|
|
},
|
|
{
|
|
"epoch": 0.6141248720573184,
|
|
"grad_norm": 0.24590979516506195,
|
|
"learning_rate": 4.074074074074074e-06,
|
|
"loss": 0.492,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.6448311156601843,
|
|
"grad_norm": 0.24135281145572662,
|
|
"learning_rate": 4.2798353909465025e-06,
|
|
"loss": 0.4861,
|
|
"step": 105
|
|
},
|
|
{
|
|
"epoch": 0.6755373592630501,
|
|
"grad_norm": 0.2440766990184784,
|
|
"learning_rate": 4.485596707818931e-06,
|
|
"loss": 0.4924,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 0.706243602865916,
|
|
"grad_norm": 0.25154179334640503,
|
|
"learning_rate": 4.691358024691358e-06,
|
|
"loss": 0.4877,
|
|
"step": 115
|
|
},
|
|
{
|
|
"epoch": 0.736949846468782,
|
|
"grad_norm": 0.2469755858182907,
|
|
"learning_rate": 4.897119341563787e-06,
|
|
"loss": 0.4821,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 0.7676560900716479,
|
|
"grad_norm": 0.282875657081604,
|
|
"learning_rate": 5.102880658436214e-06,
|
|
"loss": 0.4849,
|
|
"step": 125
|
|
},
|
|
{
|
|
"epoch": 0.7983623336745138,
|
|
"grad_norm": 0.24434781074523926,
|
|
"learning_rate": 5.3086419753086425e-06,
|
|
"loss": 0.4842,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 0.8290685772773797,
|
|
"grad_norm": 0.2680870294570923,
|
|
"learning_rate": 5.514403292181071e-06,
|
|
"loss": 0.4801,
|
|
"step": 135
|
|
},
|
|
{
|
|
"epoch": 0.8597748208802457,
|
|
"grad_norm": 0.25792691111564636,
|
|
"learning_rate": 5.720164609053498e-06,
|
|
"loss": 0.4811,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 0.8904810644831116,
|
|
"grad_norm": 0.25268977880477905,
|
|
"learning_rate": 5.925925925925926e-06,
|
|
"loss": 0.4773,
|
|
"step": 145
|
|
},
|
|
{
|
|
"epoch": 0.9211873080859775,
|
|
"grad_norm": 0.2891627848148346,
|
|
"learning_rate": 6.131687242798354e-06,
|
|
"loss": 0.48,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.9518935516888434,
|
|
"grad_norm": 0.27274736762046814,
|
|
"learning_rate": 6.3374485596707825e-06,
|
|
"loss": 0.4745,
|
|
"step": 155
|
|
},
|
|
{
|
|
"epoch": 0.9825997952917093,
|
|
"grad_norm": 0.4910643398761749,
|
|
"learning_rate": 6.543209876543211e-06,
|
|
"loss": 0.4814,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 1.0184237461617196,
|
|
"grad_norm": 0.35264864563941956,
|
|
"learning_rate": 6.748971193415639e-06,
|
|
"loss": 0.568,
|
|
"step": 165
|
|
},
|
|
{
|
|
"epoch": 1.0491299897645854,
|
|
"grad_norm": 0.2881726026535034,
|
|
"learning_rate": 6.954732510288067e-06,
|
|
"loss": 0.4704,
|
|
"step": 170
|
|
},
|
|
{
|
|
"epoch": 1.0798362333674514,
|
|
"grad_norm": 0.27516046166419983,
|
|
"learning_rate": 7.160493827160494e-06,
|
|
"loss": 0.469,
|
|
"step": 175
|
|
},
|
|
{
|
|
"epoch": 1.1105424769703174,
|
|
"grad_norm": 0.28540414571762085,
|
|
"learning_rate": 7.3662551440329225e-06,
|
|
"loss": 0.4721,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 1.1412487205731832,
|
|
"grad_norm": 0.28977257013320923,
|
|
"learning_rate": 7.57201646090535e-06,
|
|
"loss": 0.4654,
|
|
"step": 185
|
|
},
|
|
{
|
|
"epoch": 1.1719549641760492,
|
|
"grad_norm": 0.2729812264442444,
|
|
"learning_rate": 7.77777777777778e-06,
|
|
"loss": 0.4667,
|
|
"step": 190
|
|
},
|
|
{
|
|
"epoch": 1.202661207778915,
|
|
"grad_norm": 0.3296419084072113,
|
|
"learning_rate": 7.983539094650207e-06,
|
|
"loss": 0.4626,
|
|
"step": 195
|
|
},
|
|
{
|
|
"epoch": 1.233367451381781,
|
|
"grad_norm": 0.2620701491832733,
|
|
"learning_rate": 8.189300411522634e-06,
|
|
"loss": 0.4668,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 1.264073694984647,
|
|
"grad_norm": 0.29654592275619507,
|
|
"learning_rate": 8.395061728395062e-06,
|
|
"loss": 0.4618,
|
|
"step": 205
|
|
},
|
|
{
|
|
"epoch": 1.2947799385875127,
|
|
"grad_norm": 0.29271960258483887,
|
|
"learning_rate": 8.60082304526749e-06,
|
|
"loss": 0.4669,
|
|
"step": 210
|
|
},
|
|
{
|
|
"epoch": 1.3254861821903787,
|
|
"grad_norm": 0.3072218596935272,
|
|
"learning_rate": 8.806584362139918e-06,
|
|
"loss": 0.462,
|
|
"step": 215
|
|
},
|
|
{
|
|
"epoch": 1.3561924257932447,
|
|
"grad_norm": 0.28791937232017517,
|
|
"learning_rate": 9.012345679012346e-06,
|
|
"loss": 0.4644,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 1.3868986693961105,
|
|
"grad_norm": 0.49683865904808044,
|
|
"learning_rate": 9.218106995884775e-06,
|
|
"loss": 0.4644,
|
|
"step": 225
|
|
},
|
|
{
|
|
"epoch": 1.4176049129989765,
|
|
"grad_norm": 0.3112960159778595,
|
|
"learning_rate": 9.423868312757202e-06,
|
|
"loss": 0.4658,
|
|
"step": 230
|
|
},
|
|
{
|
|
"epoch": 1.4483111566018425,
|
|
"grad_norm": 0.2987211048603058,
|
|
"learning_rate": 9.62962962962963e-06,
|
|
"loss": 0.457,
|
|
"step": 235
|
|
},
|
|
{
|
|
"epoch": 1.4790174002047083,
|
|
"grad_norm": 0.32146817445755005,
|
|
"learning_rate": 9.835390946502057e-06,
|
|
"loss": 0.4617,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 1.509723643807574,
|
|
"grad_norm": 0.5294284820556641,
|
|
"learning_rate": 9.999994841278135e-06,
|
|
"loss": 0.4639,
|
|
"step": 245
|
|
},
|
|
{
|
|
"epoch": 1.54042988741044,
|
|
"grad_norm": 0.28634026646614075,
|
|
"learning_rate": 9.99981428713058e-06,
|
|
"loss": 0.4641,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 1.571136131013306,
|
|
"grad_norm": 0.30058643221855164,
|
|
"learning_rate": 9.999375807534642e-06,
|
|
"loss": 0.4614,
|
|
"step": 255
|
|
},
|
|
{
|
|
"epoch": 1.6018423746161718,
|
|
"grad_norm": 0.2915442883968353,
|
|
"learning_rate": 9.998679425110168e-06,
|
|
"loss": 0.4627,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 1.6325486182190379,
|
|
"grad_norm": 0.3363260328769684,
|
|
"learning_rate": 9.997725175781445e-06,
|
|
"loss": 0.4609,
|
|
"step": 265
|
|
},
|
|
{
|
|
"epoch": 1.6632548618219039,
|
|
"grad_norm": 0.31282299757003784,
|
|
"learning_rate": 9.996513108775338e-06,
|
|
"loss": 0.4591,
|
|
"step": 270
|
|
},
|
|
{
|
|
"epoch": 1.6939611054247696,
|
|
"grad_norm": 0.3025549650192261,
|
|
"learning_rate": 9.995043286618752e-06,
|
|
"loss": 0.4593,
|
|
"step": 275
|
|
},
|
|
{
|
|
"epoch": 1.7246673490276356,
|
|
"grad_norm": 0.2985345721244812,
|
|
"learning_rate": 9.993315785135417e-06,
|
|
"loss": 0.4576,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 1.7553735926305016,
|
|
"grad_norm": 0.3218015134334564,
|
|
"learning_rate": 9.991330693441956e-06,
|
|
"loss": 0.4553,
|
|
"step": 285
|
|
},
|
|
{
|
|
"epoch": 1.7860798362333674,
|
|
"grad_norm": 0.32680895924568176,
|
|
"learning_rate": 9.989088113943309e-06,
|
|
"loss": 0.4583,
|
|
"step": 290
|
|
},
|
|
{
|
|
"epoch": 1.8167860798362334,
|
|
"grad_norm": 0.29413729906082153,
|
|
"learning_rate": 9.986588162327436e-06,
|
|
"loss": 0.4598,
|
|
"step": 295
|
|
},
|
|
{
|
|
"epoch": 1.8474923234390994,
|
|
"grad_norm": 0.3157609701156616,
|
|
"learning_rate": 9.983830967559355e-06,
|
|
"loss": 0.4566,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 1.8781985670419652,
|
|
"grad_norm": 0.29272714257240295,
|
|
"learning_rate": 9.98081667187449e-06,
|
|
"loss": 0.4521,
|
|
"step": 305
|
|
},
|
|
{
|
|
"epoch": 1.908904810644831,
|
|
"grad_norm": 0.31150153279304504,
|
|
"learning_rate": 9.977545430771332e-06,
|
|
"loss": 0.4582,
|
|
"step": 310
|
|
},
|
|
{
|
|
"epoch": 1.9396110542476972,
|
|
"grad_norm": 0.31409797072410583,
|
|
"learning_rate": 9.974017413003407e-06,
|
|
"loss": 0.4543,
|
|
"step": 315
|
|
},
|
|
{
|
|
"epoch": 1.970317297850563,
|
|
"grad_norm": 0.3109057545661926,
|
|
"learning_rate": 9.970232800570594e-06,
|
|
"loss": 0.4543,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 2.006141248720573,
|
|
"grad_norm": 0.656970202922821,
|
|
"learning_rate": 9.966191788709716e-06,
|
|
"loss": 0.5494,
|
|
"step": 325
|
|
},
|
|
{
|
|
"epoch": 2.036847492323439,
|
|
"grad_norm": 0.3206768333911896,
|
|
"learning_rate": 9.961894585884472e-06,
|
|
"loss": 0.4347,
|
|
"step": 330
|
|
},
|
|
{
|
|
"epoch": 2.067553735926305,
|
|
"grad_norm": 0.35229599475860596,
|
|
"learning_rate": 9.957341413774693e-06,
|
|
"loss": 0.4389,
|
|
"step": 335
|
|
},
|
|
{
|
|
"epoch": 2.0982599795291708,
|
|
"grad_norm": 0.3107539117336273,
|
|
"learning_rate": 9.952532507264892e-06,
|
|
"loss": 0.4381,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 2.128966223132037,
|
|
"grad_norm": 0.3630954325199127,
|
|
"learning_rate": 9.947468114432156e-06,
|
|
"loss": 0.4396,
|
|
"step": 345
|
|
},
|
|
{
|
|
"epoch": 2.1596724667349028,
|
|
"grad_norm": 0.3643966019153595,
|
|
"learning_rate": 9.942148496533348e-06,
|
|
"loss": 0.4403,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 2.1903787103377685,
|
|
"grad_norm": 0.2925087809562683,
|
|
"learning_rate": 9.936573927991631e-06,
|
|
"loss": 0.4381,
|
|
"step": 355
|
|
},
|
|
{
|
|
"epoch": 2.2210849539406348,
|
|
"grad_norm": 0.30517083406448364,
|
|
"learning_rate": 9.930744696382298e-06,
|
|
"loss": 0.4372,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 2.2517911975435005,
|
|
"grad_norm": 0.30064260959625244,
|
|
"learning_rate": 9.924661102417959e-06,
|
|
"loss": 0.4413,
|
|
"step": 365
|
|
},
|
|
{
|
|
"epoch": 2.2824974411463663,
|
|
"grad_norm": 0.3043610155582428,
|
|
"learning_rate": 9.918323459933006e-06,
|
|
"loss": 0.4373,
|
|
"step": 370
|
|
},
|
|
{
|
|
"epoch": 2.313203684749232,
|
|
"grad_norm": 0.35172680020332336,
|
|
"learning_rate": 9.911732095867443e-06,
|
|
"loss": 0.4369,
|
|
"step": 375
|
|
},
|
|
{
|
|
"epoch": 2.3439099283520983,
|
|
"grad_norm": 0.3194226026535034,
|
|
"learning_rate": 9.904887350250002e-06,
|
|
"loss": 0.4377,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 2.374616171954964,
|
|
"grad_norm": 0.33400508761405945,
|
|
"learning_rate": 9.897789576180617e-06,
|
|
"loss": 0.4389,
|
|
"step": 385
|
|
},
|
|
{
|
|
"epoch": 2.40532241555783,
|
|
"grad_norm": 0.3146711587905884,
|
|
"learning_rate": 9.8904391398122e-06,
|
|
"loss": 0.4392,
|
|
"step": 390
|
|
},
|
|
{
|
|
"epoch": 2.436028659160696,
|
|
"grad_norm": 0.3004317581653595,
|
|
"learning_rate": 9.882836420331753e-06,
|
|
"loss": 0.4382,
|
|
"step": 395
|
|
},
|
|
{
|
|
"epoch": 2.466734902763562,
|
|
"grad_norm": 0.28334906697273254,
|
|
"learning_rate": 9.87498180994081e-06,
|
|
"loss": 0.4369,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 2.4974411463664277,
|
|
"grad_norm": 0.31731870770454407,
|
|
"learning_rate": 9.8668757138352e-06,
|
|
"loss": 0.4371,
|
|
"step": 405
|
|
},
|
|
{
|
|
"epoch": 2.528147389969294,
|
|
"grad_norm": 0.31999263167381287,
|
|
"learning_rate": 9.858518550184154e-06,
|
|
"loss": 0.4361,
|
|
"step": 410
|
|
},
|
|
{
|
|
"epoch": 2.5588536335721597,
|
|
"grad_norm": 0.288147896528244,
|
|
"learning_rate": 9.849910750108718e-06,
|
|
"loss": 0.438,
|
|
"step": 415
|
|
},
|
|
{
|
|
"epoch": 2.5895598771750254,
|
|
"grad_norm": 0.28145623207092285,
|
|
"learning_rate": 9.841052757659525e-06,
|
|
"loss": 0.4375,
|
|
"step": 420
|
|
},
|
|
{
|
|
"epoch": 2.6202661207778917,
|
|
"grad_norm": 0.30709531903266907,
|
|
"learning_rate": 9.831945029793884e-06,
|
|
"loss": 0.4403,
|
|
"step": 425
|
|
},
|
|
{
|
|
"epoch": 2.6509723643807575,
|
|
"grad_norm": 0.285591185092926,
|
|
"learning_rate": 9.822588036352201e-06,
|
|
"loss": 0.4375,
|
|
"step": 430
|
|
},
|
|
{
|
|
"epoch": 2.6816786079836232,
|
|
"grad_norm": 0.2885895371437073,
|
|
"learning_rate": 9.812982260033753e-06,
|
|
"loss": 0.4402,
|
|
"step": 435
|
|
},
|
|
{
|
|
"epoch": 2.7123848515864895,
|
|
"grad_norm": 0.28683528304100037,
|
|
"learning_rate": 9.803128196371778e-06,
|
|
"loss": 0.4364,
|
|
"step": 440
|
|
},
|
|
{
|
|
"epoch": 2.7430910951893552,
|
|
"grad_norm": 0.3029756247997284,
|
|
"learning_rate": 9.793026353707915e-06,
|
|
"loss": 0.4377,
|
|
"step": 445
|
|
},
|
|
{
|
|
"epoch": 2.773797338792221,
|
|
"grad_norm": 0.3118147850036621,
|
|
"learning_rate": 9.782677253165979e-06,
|
|
"loss": 0.4344,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 2.8045035823950872,
|
|
"grad_norm": 0.30943065881729126,
|
|
"learning_rate": 9.77208142862508e-06,
|
|
"loss": 0.4377,
|
|
"step": 455
|
|
},
|
|
{
|
|
"epoch": 2.835209825997953,
|
|
"grad_norm": 0.3110142946243286,
|
|
"learning_rate": 9.761239426692077e-06,
|
|
"loss": 0.4361,
|
|
"step": 460
|
|
},
|
|
{
|
|
"epoch": 2.865916069600819,
|
|
"grad_norm": 0.29598164558410645,
|
|
"learning_rate": 9.750151806673389e-06,
|
|
"loss": 0.4357,
|
|
"step": 465
|
|
},
|
|
{
|
|
"epoch": 2.896622313203685,
|
|
"grad_norm": 0.3019062280654907,
|
|
"learning_rate": 9.738819140546135e-06,
|
|
"loss": 0.433,
|
|
"step": 470
|
|
},
|
|
{
|
|
"epoch": 2.927328556806551,
|
|
"grad_norm": 0.28945985436439514,
|
|
"learning_rate": 9.727242012928622e-06,
|
|
"loss": 0.4397,
|
|
"step": 475
|
|
},
|
|
{
|
|
"epoch": 2.9580348004094166,
|
|
"grad_norm": 0.31174397468566895,
|
|
"learning_rate": 9.715421021050205e-06,
|
|
"loss": 0.4331,
|
|
"step": 480
|
|
},
|
|
{
|
|
"epoch": 2.9887410440122824,
|
|
"grad_norm": 0.306142121553421,
|
|
"learning_rate": 9.703356774720454e-06,
|
|
"loss": 0.4356,
|
|
"step": 485
|
|
},
|
|
{
|
|
"epoch": 3.024564994882293,
|
|
"grad_norm": 0.3875485062599182,
|
|
"learning_rate": 9.69104989629772e-06,
|
|
"loss": 0.5058,
|
|
"step": 490
|
|
},
|
|
{
|
|
"epoch": 3.0552712384851586,
|
|
"grad_norm": 0.3581351339817047,
|
|
"learning_rate": 9.678501020657008e-06,
|
|
"loss": 0.4168,
|
|
"step": 495
|
|
},
|
|
{
|
|
"epoch": 3.0859774820880244,
|
|
"grad_norm": 0.32925835251808167,
|
|
"learning_rate": 9.665710795157236e-06,
|
|
"loss": 0.4162,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 3.1166837256908906,
|
|
"grad_norm": 0.33935922384262085,
|
|
"learning_rate": 9.652679879607843e-06,
|
|
"loss": 0.4188,
|
|
"step": 505
|
|
},
|
|
{
|
|
"epoch": 3.1473899692937564,
|
|
"grad_norm": 0.3029758930206299,
|
|
"learning_rate": 9.639408946234745e-06,
|
|
"loss": 0.4175,
|
|
"step": 510
|
|
},
|
|
{
|
|
"epoch": 3.178096212896622,
|
|
"grad_norm": 0.32193148136138916,
|
|
"learning_rate": 9.625898679645656e-06,
|
|
"loss": 0.4163,
|
|
"step": 515
|
|
},
|
|
{
|
|
"epoch": 3.2088024564994884,
|
|
"grad_norm": 0.34081465005874634,
|
|
"learning_rate": 9.612149776794776e-06,
|
|
"loss": 0.4153,
|
|
"step": 520
|
|
},
|
|
{
|
|
"epoch": 3.239508700102354,
|
|
"grad_norm": 0.31485581398010254,
|
|
"learning_rate": 9.59816294694684e-06,
|
|
"loss": 0.4133,
|
|
"step": 525
|
|
},
|
|
{
|
|
"epoch": 3.27021494370522,
|
|
"grad_norm": 0.32688620686531067,
|
|
"learning_rate": 9.583938911640513e-06,
|
|
"loss": 0.4211,
|
|
"step": 530
|
|
},
|
|
{
|
|
"epoch": 3.300921187308086,
|
|
"grad_norm": 0.3346046209335327,
|
|
"learning_rate": 9.569478404651192e-06,
|
|
"loss": 0.4172,
|
|
"step": 535
|
|
},
|
|
{
|
|
"epoch": 3.331627430910952,
|
|
"grad_norm": 0.29652711749076843,
|
|
"learning_rate": 9.55478217195313e-06,
|
|
"loss": 0.4146,
|
|
"step": 540
|
|
},
|
|
{
|
|
"epoch": 3.3623336745138177,
|
|
"grad_norm": 0.33152851462364197,
|
|
"learning_rate": 9.53985097168097e-06,
|
|
"loss": 0.4185,
|
|
"step": 545
|
|
},
|
|
{
|
|
"epoch": 3.393039918116684,
|
|
"grad_norm": 0.32041046023368835,
|
|
"learning_rate": 9.524685574090627e-06,
|
|
"loss": 0.4178,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 3.4237461617195497,
|
|
"grad_norm": 0.3175600469112396,
|
|
"learning_rate": 9.50928676151955e-06,
|
|
"loss": 0.418,
|
|
"step": 555
|
|
},
|
|
{
|
|
"epoch": 3.4544524053224155,
|
|
"grad_norm": 0.4362170696258545,
|
|
"learning_rate": 9.493655328346378e-06,
|
|
"loss": 0.4181,
|
|
"step": 560
|
|
},
|
|
{
|
|
"epoch": 3.4851586489252817,
|
|
"grad_norm": 0.28091391921043396,
|
|
"learning_rate": 9.477792080949938e-06,
|
|
"loss": 0.4193,
|
|
"step": 565
|
|
},
|
|
{
|
|
"epoch": 3.5158648925281475,
|
|
"grad_norm": 0.30598944425582886,
|
|
"learning_rate": 9.461697837667668e-06,
|
|
"loss": 0.418,
|
|
"step": 570
|
|
},
|
|
{
|
|
"epoch": 3.5465711361310133,
|
|
"grad_norm": 0.2786954641342163,
|
|
"learning_rate": 9.445373428753386e-06,
|
|
"loss": 0.4176,
|
|
"step": 575
|
|
},
|
|
{
|
|
"epoch": 3.5772773797338795,
|
|
"grad_norm": 0.30698952078819275,
|
|
"learning_rate": 9.42881969633447e-06,
|
|
"loss": 0.4171,
|
|
"step": 580
|
|
},
|
|
{
|
|
"epoch": 3.6079836233367453,
|
|
"grad_norm": 0.3353785276412964,
|
|
"learning_rate": 9.412037494368412e-06,
|
|
"loss": 0.4212,
|
|
"step": 585
|
|
},
|
|
{
|
|
"epoch": 3.638689866939611,
|
|
"grad_norm": 0.48117902874946594,
|
|
"learning_rate": 9.395027688598756e-06,
|
|
"loss": 0.4201,
|
|
"step": 590
|
|
},
|
|
{
|
|
"epoch": 3.669396110542477,
|
|
"grad_norm": 0.2947028577327728,
|
|
"learning_rate": 9.377791156510456e-06,
|
|
"loss": 0.4211,
|
|
"step": 595
|
|
},
|
|
{
|
|
"epoch": 3.7001023541453426,
|
|
"grad_norm": 0.3083222210407257,
|
|
"learning_rate": 9.360328787284587e-06,
|
|
"loss": 0.4184,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 3.730808597748209,
|
|
"grad_norm": 0.28462743759155273,
|
|
"learning_rate": 9.342641481752492e-06,
|
|
"loss": 0.4194,
|
|
"step": 605
|
|
},
|
|
{
|
|
"epoch": 3.7615148413510746,
|
|
"grad_norm": 0.29272404313087463,
|
|
"learning_rate": 9.324730152349305e-06,
|
|
"loss": 0.4139,
|
|
"step": 610
|
|
},
|
|
{
|
|
"epoch": 3.7922210849539404,
|
|
"grad_norm": 0.305203914642334,
|
|
"learning_rate": 9.306595723066878e-06,
|
|
"loss": 0.4205,
|
|
"step": 615
|
|
},
|
|
{
|
|
"epoch": 3.8229273285568066,
|
|
"grad_norm": 0.3017217516899109,
|
|
"learning_rate": 9.28823912940612e-06,
|
|
"loss": 0.4195,
|
|
"step": 620
|
|
},
|
|
{
|
|
"epoch": 3.8536335721596724,
|
|
"grad_norm": 0.30164411664009094,
|
|
"learning_rate": 9.26966131832873e-06,
|
|
"loss": 0.4201,
|
|
"step": 625
|
|
},
|
|
{
|
|
"epoch": 3.884339815762538,
|
|
"grad_norm": 0.2872079312801361,
|
|
"learning_rate": 9.250863248208357e-06,
|
|
"loss": 0.4209,
|
|
"step": 630
|
|
},
|
|
{
|
|
"epoch": 3.9150460593654044,
|
|
"grad_norm": 0.29657259583473206,
|
|
"learning_rate": 9.231845888781153e-06,
|
|
"loss": 0.4179,
|
|
"step": 635
|
|
},
|
|
{
|
|
"epoch": 3.94575230296827,
|
|
"grad_norm": 0.2927268147468567,
|
|
"learning_rate": 9.212610221095748e-06,
|
|
"loss": 0.421,
|
|
"step": 640
|
|
},
|
|
{
|
|
"epoch": 3.976458546571136,
|
|
"grad_norm": 0.37469282746315,
|
|
"learning_rate": 9.193157237462642e-06,
|
|
"loss": 0.4188,
|
|
"step": 645
|
|
},
|
|
{
|
|
"epoch": 4.012282497441146,
|
|
"grad_norm": 0.41009268164634705,
|
|
"learning_rate": 9.173487941403011e-06,
|
|
"loss": 0.4919,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 4.042988741044012,
|
|
"grad_norm": 0.3871892988681793,
|
|
"learning_rate": 9.153603347596946e-06,
|
|
"loss": 0.3978,
|
|
"step": 655
|
|
},
|
|
{
|
|
"epoch": 4.073694984646878,
|
|
"grad_norm": 0.34029635787010193,
|
|
"learning_rate": 9.133504481831103e-06,
|
|
"loss": 0.4014,
|
|
"step": 660
|
|
},
|
|
{
|
|
"epoch": 4.104401228249744,
|
|
"grad_norm": 0.3581830859184265,
|
|
"learning_rate": 9.113192380945783e-06,
|
|
"loss": 0.3992,
|
|
"step": 665
|
|
},
|
|
{
|
|
"epoch": 4.13510747185261,
|
|
"grad_norm": 0.33405575156211853,
|
|
"learning_rate": 9.092668092781454e-06,
|
|
"loss": 0.4003,
|
|
"step": 670
|
|
},
|
|
{
|
|
"epoch": 4.165813715455476,
|
|
"grad_norm": 0.35374191403388977,
|
|
"learning_rate": 9.071932676124686e-06,
|
|
"loss": 0.3983,
|
|
"step": 675
|
|
},
|
|
{
|
|
"epoch": 4.1965199590583415,
|
|
"grad_norm": 0.32420825958251953,
|
|
"learning_rate": 9.050987200653538e-06,
|
|
"loss": 0.3992,
|
|
"step": 680
|
|
},
|
|
{
|
|
"epoch": 4.227226202661208,
|
|
"grad_norm": 0.307517409324646,
|
|
"learning_rate": 9.029832746882372e-06,
|
|
"loss": 0.4005,
|
|
"step": 685
|
|
},
|
|
{
|
|
"epoch": 4.257932446264074,
|
|
"grad_norm": 0.3153163492679596,
|
|
"learning_rate": 9.008470406106118e-06,
|
|
"loss": 0.3957,
|
|
"step": 690
|
|
},
|
|
{
|
|
"epoch": 4.288638689866939,
|
|
"grad_norm": 0.32640504837036133,
|
|
"learning_rate": 8.986901280343973e-06,
|
|
"loss": 0.4002,
|
|
"step": 695
|
|
},
|
|
{
|
|
"epoch": 4.3193449334698055,
|
|
"grad_norm": 0.29307159781455994,
|
|
"learning_rate": 8.96512648228255e-06,
|
|
"loss": 0.4008,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 4.350051177072672,
|
|
"grad_norm": 0.3047451078891754,
|
|
"learning_rate": 8.943147135218482e-06,
|
|
"loss": 0.4006,
|
|
"step": 705
|
|
},
|
|
{
|
|
"epoch": 4.380757420675537,
|
|
"grad_norm": 0.3649887442588806,
|
|
"learning_rate": 8.920964373000474e-06,
|
|
"loss": 0.4013,
|
|
"step": 710
|
|
},
|
|
{
|
|
"epoch": 4.411463664278403,
|
|
"grad_norm": 0.630779504776001,
|
|
"learning_rate": 8.898579339970806e-06,
|
|
"loss": 0.4024,
|
|
"step": 715
|
|
},
|
|
{
|
|
"epoch": 4.4421699078812695,
|
|
"grad_norm": 0.2852974236011505,
|
|
"learning_rate": 8.875993190906309e-06,
|
|
"loss": 0.4028,
|
|
"step": 720
|
|
},
|
|
{
|
|
"epoch": 4.472876151484135,
|
|
"grad_norm": 0.35749351978302,
|
|
"learning_rate": 8.85320709095878e-06,
|
|
"loss": 0.4001,
|
|
"step": 725
|
|
},
|
|
{
|
|
"epoch": 4.503582395087001,
|
|
"grad_norm": 0.32927122712135315,
|
|
"learning_rate": 8.83022221559489e-06,
|
|
"loss": 0.4007,
|
|
"step": 730
|
|
},
|
|
{
|
|
"epoch": 4.534288638689867,
|
|
"grad_norm": 0.2845518887042999,
|
|
"learning_rate": 8.80703975053554e-06,
|
|
"loss": 0.3998,
|
|
"step": 735
|
|
},
|
|
{
|
|
"epoch": 4.564994882292733,
|
|
"grad_norm": 0.29234543442726135,
|
|
"learning_rate": 8.783660891694683e-06,
|
|
"loss": 0.4031,
|
|
"step": 740
|
|
},
|
|
{
|
|
"epoch": 4.595701125895599,
|
|
"grad_norm": 0.2943549156188965,
|
|
"learning_rate": 8.760086845117648e-06,
|
|
"loss": 0.3995,
|
|
"step": 745
|
|
},
|
|
{
|
|
"epoch": 4.626407369498464,
|
|
"grad_norm": 0.2838557958602905,
|
|
"learning_rate": 8.736318826918909e-06,
|
|
"loss": 0.3998,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 4.65711361310133,
|
|
"grad_norm": 0.29005348682403564,
|
|
"learning_rate": 8.71235806321936e-06,
|
|
"loss": 0.4007,
|
|
"step": 755
|
|
},
|
|
{
|
|
"epoch": 4.687819856704197,
|
|
"grad_norm": 0.300619512796402,
|
|
"learning_rate": 8.688205790083053e-06,
|
|
"loss": 0.3994,
|
|
"step": 760
|
|
},
|
|
{
|
|
"epoch": 4.718526100307063,
|
|
"grad_norm": 0.3142443895339966,
|
|
"learning_rate": 8.663863253453444e-06,
|
|
"loss": 0.4026,
|
|
"step": 765
|
|
},
|
|
{
|
|
"epoch": 4.749232343909928,
|
|
"grad_norm": 0.3032344877719879,
|
|
"learning_rate": 8.639331709089107e-06,
|
|
"loss": 0.403,
|
|
"step": 770
|
|
},
|
|
{
|
|
"epoch": 4.779938587512794,
|
|
"grad_norm": 0.4341191351413727,
|
|
"learning_rate": 8.614612422498965e-06,
|
|
"loss": 0.4014,
|
|
"step": 775
|
|
},
|
|
{
|
|
"epoch": 4.81064483111566,
|
|
"grad_norm": 0.29899492859840393,
|
|
"learning_rate": 8.589706668876995e-06,
|
|
"loss": 0.4029,
|
|
"step": 780
|
|
},
|
|
{
|
|
"epoch": 4.841351074718526,
|
|
"grad_norm": 0.30844879150390625,
|
|
"learning_rate": 8.564615733036457e-06,
|
|
"loss": 0.4009,
|
|
"step": 785
|
|
},
|
|
{
|
|
"epoch": 4.872057318321392,
|
|
"grad_norm": 0.3283848166465759,
|
|
"learning_rate": 8.539340909343597e-06,
|
|
"loss": 0.4042,
|
|
"step": 790
|
|
},
|
|
{
|
|
"epoch": 4.9027635619242576,
|
|
"grad_norm": 0.287864089012146,
|
|
"learning_rate": 8.513883501650892e-06,
|
|
"loss": 0.403,
|
|
"step": 795
|
|
},
|
|
{
|
|
"epoch": 4.933469805527124,
|
|
"grad_norm": 0.3053203523159027,
|
|
"learning_rate": 8.488244823229781e-06,
|
|
"loss": 0.4041,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 4.96417604912999,
|
|
"grad_norm": 0.30933913588523865,
|
|
"learning_rate": 8.462426196702912e-06,
|
|
"loss": 0.3996,
|
|
"step": 805
|
|
},
|
|
{
|
|
"epoch": 4.994882292732855,
|
|
"grad_norm": 0.29554951190948486,
|
|
"learning_rate": 8.436428953975921e-06,
|
|
"loss": 0.3998,
|
|
"step": 810
|
|
},
|
|
{
|
|
"epoch": 5.030706243602866,
|
|
"grad_norm": 0.38984817266464233,
|
|
"learning_rate": 8.41025443616872e-06,
|
|
"loss": 0.4632,
|
|
"step": 815
|
|
},
|
|
{
|
|
"epoch": 5.061412487205732,
|
|
"grad_norm": 0.33434659242630005,
|
|
"learning_rate": 8.38390399354631e-06,
|
|
"loss": 0.3828,
|
|
"step": 820
|
|
},
|
|
{
|
|
"epoch": 5.092118730808598,
|
|
"grad_norm": 0.3267410099506378,
|
|
"learning_rate": 8.357378985449124e-06,
|
|
"loss": 0.382,
|
|
"step": 825
|
|
},
|
|
{
|
|
"epoch": 5.122824974411464,
|
|
"grad_norm": 0.33452486991882324,
|
|
"learning_rate": 8.330680780222907e-06,
|
|
"loss": 0.3821,
|
|
"step": 830
|
|
},
|
|
{
|
|
"epoch": 5.153531218014329,
|
|
"grad_norm": 0.3105202913284302,
|
|
"learning_rate": 8.303810755148127e-06,
|
|
"loss": 0.3843,
|
|
"step": 835
|
|
},
|
|
{
|
|
"epoch": 5.184237461617196,
|
|
"grad_norm": 0.3055059313774109,
|
|
"learning_rate": 8.276770296368922e-06,
|
|
"loss": 0.3873,
|
|
"step": 840
|
|
},
|
|
{
|
|
"epoch": 5.214943705220062,
|
|
"grad_norm": 0.2939026653766632,
|
|
"learning_rate": 8.249560798821592e-06,
|
|
"loss": 0.3822,
|
|
"step": 845
|
|
},
|
|
{
|
|
"epoch": 5.245649948822927,
|
|
"grad_norm": 0.32671213150024414,
|
|
"learning_rate": 8.222183666162647e-06,
|
|
"loss": 0.3825,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 5.276356192425793,
|
|
"grad_norm": 0.3374998867511749,
|
|
"learning_rate": 8.194640310696383e-06,
|
|
"loss": 0.3865,
|
|
"step": 855
|
|
},
|
|
{
|
|
"epoch": 5.30706243602866,
|
|
"grad_norm": 0.3487096130847931,
|
|
"learning_rate": 8.16693215330204e-06,
|
|
"loss": 0.3833,
|
|
"step": 860
|
|
},
|
|
{
|
|
"epoch": 5.337768679631525,
|
|
"grad_norm": 0.3201228976249695,
|
|
"learning_rate": 8.139060623360494e-06,
|
|
"loss": 0.3837,
|
|
"step": 865
|
|
},
|
|
{
|
|
"epoch": 5.368474923234391,
|
|
"grad_norm": 0.30659347772598267,
|
|
"learning_rate": 8.111027158680516e-06,
|
|
"loss": 0.3842,
|
|
"step": 870
|
|
},
|
|
{
|
|
"epoch": 5.399181166837257,
|
|
"grad_norm": 0.2944610118865967,
|
|
"learning_rate": 8.082833205424614e-06,
|
|
"loss": 0.385,
|
|
"step": 875
|
|
},
|
|
{
|
|
"epoch": 5.429887410440123,
|
|
"grad_norm": 0.3064741790294647,
|
|
"learning_rate": 8.054480218034415e-06,
|
|
"loss": 0.3836,
|
|
"step": 880
|
|
},
|
|
{
|
|
"epoch": 5.460593654042989,
|
|
"grad_norm": 0.3105819523334503,
|
|
"learning_rate": 8.02596965915564e-06,
|
|
"loss": 0.3864,
|
|
"step": 885
|
|
},
|
|
{
|
|
"epoch": 5.491299897645854,
|
|
"grad_norm": 0.3280338644981384,
|
|
"learning_rate": 7.997302999562657e-06,
|
|
"loss": 0.3836,
|
|
"step": 890
|
|
},
|
|
{
|
|
"epoch": 5.5220061412487205,
|
|
"grad_norm": 0.3340366780757904,
|
|
"learning_rate": 7.968481718082601e-06,
|
|
"loss": 0.382,
|
|
"step": 895
|
|
},
|
|
{
|
|
"epoch": 5.552712384851587,
|
|
"grad_norm": 0.30426710844039917,
|
|
"learning_rate": 7.93950730151908e-06,
|
|
"loss": 0.3825,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 5.583418628454452,
|
|
"grad_norm": 0.3064902126789093,
|
|
"learning_rate": 7.910381244575491e-06,
|
|
"loss": 0.3838,
|
|
"step": 905
|
|
},
|
|
{
|
|
"epoch": 5.614124872057318,
|
|
"grad_norm": 0.2979416251182556,
|
|
"learning_rate": 7.881105049777902e-06,
|
|
"loss": 0.3876,
|
|
"step": 910
|
|
},
|
|
{
|
|
"epoch": 5.6448311156601845,
|
|
"grad_norm": 0.29067227244377136,
|
|
"learning_rate": 7.851680227397541e-06,
|
|
"loss": 0.3898,
|
|
"step": 915
|
|
},
|
|
{
|
|
"epoch": 5.67553735926305,
|
|
"grad_norm": 0.3097393214702606,
|
|
"learning_rate": 7.82210829537289e-06,
|
|
"loss": 0.3821,
|
|
"step": 920
|
|
},
|
|
{
|
|
"epoch": 5.706243602865916,
|
|
"grad_norm": 0.3296332061290741,
|
|
"learning_rate": 7.792390779231374e-06,
|
|
"loss": 0.3853,
|
|
"step": 925
|
|
},
|
|
{
|
|
"epoch": 5.736949846468782,
|
|
"grad_norm": 0.3411824405193329,
|
|
"learning_rate": 7.762529212010675e-06,
|
|
"loss": 0.3857,
|
|
"step": 930
|
|
},
|
|
{
|
|
"epoch": 5.767656090071648,
|
|
"grad_norm": 0.30363622307777405,
|
|
"learning_rate": 7.732525134179625e-06,
|
|
"loss": 0.3869,
|
|
"step": 935
|
|
},
|
|
{
|
|
"epoch": 5.798362333674514,
|
|
"grad_norm": 0.38818359375,
|
|
"learning_rate": 7.702380093558766e-06,
|
|
"loss": 0.3884,
|
|
"step": 940
|
|
},
|
|
{
|
|
"epoch": 5.82906857727738,
|
|
"grad_norm": 0.302729070186615,
|
|
"learning_rate": 7.672095645240479e-06,
|
|
"loss": 0.3849,
|
|
"step": 945
|
|
},
|
|
{
|
|
"epoch": 5.859774820880245,
|
|
"grad_norm": 0.29930469393730164,
|
|
"learning_rate": 7.641673351508774e-06,
|
|
"loss": 0.3835,
|
|
"step": 950
|
|
},
|
|
{
|
|
"epoch": 5.890481064483112,
|
|
"grad_norm": 0.28438737988471985,
|
|
"learning_rate": 7.6111147817586925e-06,
|
|
"loss": 0.385,
|
|
"step": 955
|
|
},
|
|
{
|
|
"epoch": 5.921187308085978,
|
|
"grad_norm": 0.2847469747066498,
|
|
"learning_rate": 7.580421512415349e-06,
|
|
"loss": 0.385,
|
|
"step": 960
|
|
},
|
|
{
|
|
"epoch": 5.951893551688843,
|
|
"grad_norm": 0.2834565043449402,
|
|
"learning_rate": 7.549595126852605e-06,
|
|
"loss": 0.3846,
|
|
"step": 965
|
|
},
|
|
{
|
|
"epoch": 5.982599795291709,
|
|
"grad_norm": 0.2991902232170105,
|
|
"learning_rate": 7.518637215311388e-06,
|
|
"loss": 0.385,
|
|
"step": 970
|
|
},
|
|
{
|
|
"epoch": 6.018423746161719,
|
|
"grad_norm": 0.3554234802722931,
|
|
"learning_rate": 7.487549374817662e-06,
|
|
"loss": 0.4497,
|
|
"step": 975
|
|
},
|
|
{
|
|
"epoch": 6.049129989764586,
|
|
"grad_norm": 0.379526287317276,
|
|
"learning_rate": 7.456333209100032e-06,
|
|
"loss": 0.367,
|
|
"step": 980
|
|
},
|
|
{
|
|
"epoch": 6.079836233367452,
|
|
"grad_norm": 0.32687127590179443,
|
|
"learning_rate": 7.424990328507017e-06,
|
|
"loss": 0.3642,
|
|
"step": 985
|
|
},
|
|
{
|
|
"epoch": 6.110542476970317,
|
|
"grad_norm": 0.3256913423538208,
|
|
"learning_rate": 7.393522349923981e-06,
|
|
"loss": 0.3662,
|
|
"step": 990
|
|
},
|
|
{
|
|
"epoch": 6.141248720573183,
|
|
"grad_norm": 0.3289490044116974,
|
|
"learning_rate": 7.361930896689713e-06,
|
|
"loss": 0.3668,
|
|
"step": 995
|
|
},
|
|
{
|
|
"epoch": 6.171954964176049,
|
|
"grad_norm": 0.3317910134792328,
|
|
"learning_rate": 7.330217598512696e-06,
|
|
"loss": 0.3698,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 6.202661207778915,
|
|
"grad_norm": 0.3371947705745697,
|
|
"learning_rate": 7.2983840913870215e-06,
|
|
"loss": 0.3669,
|
|
"step": 1005
|
|
},
|
|
{
|
|
"epoch": 6.233367451381781,
|
|
"grad_norm": 0.32799994945526123,
|
|
"learning_rate": 7.266432017508008e-06,
|
|
"loss": 0.3699,
|
|
"step": 1010
|
|
},
|
|
{
|
|
"epoch": 6.2640736949846465,
|
|
"grad_norm": 0.3218367397785187,
|
|
"learning_rate": 7.234363025187474e-06,
|
|
"loss": 0.3661,
|
|
"step": 1015
|
|
},
|
|
{
|
|
"epoch": 6.294779938587513,
|
|
"grad_norm": 0.3230588436126709,
|
|
"learning_rate": 7.202178768768711e-06,
|
|
"loss": 0.3663,
|
|
"step": 1020
|
|
},
|
|
{
|
|
"epoch": 6.325486182190379,
|
|
"grad_norm": 0.3200216591358185,
|
|
"learning_rate": 7.169880908541136e-06,
|
|
"loss": 0.3654,
|
|
"step": 1025
|
|
},
|
|
{
|
|
"epoch": 6.356192425793244,
|
|
"grad_norm": 0.3350008726119995,
|
|
"learning_rate": 7.137471110654656e-06,
|
|
"loss": 0.3676,
|
|
"step": 1030
|
|
},
|
|
{
|
|
"epoch": 6.3868986693961105,
|
|
"grad_norm": 0.32792818546295166,
|
|
"learning_rate": 7.104951047033697e-06,
|
|
"loss": 0.3666,
|
|
"step": 1035
|
|
},
|
|
{
|
|
"epoch": 6.417604912998977,
|
|
"grad_norm": 0.31437090039253235,
|
|
"learning_rate": 7.0723223952909694e-06,
|
|
"loss": 0.3668,
|
|
"step": 1040
|
|
},
|
|
{
|
|
"epoch": 6.448311156601842,
|
|
"grad_norm": 0.30968931317329407,
|
|
"learning_rate": 7.039586838640918e-06,
|
|
"loss": 0.367,
|
|
"step": 1045
|
|
},
|
|
{
|
|
"epoch": 6.479017400204708,
|
|
"grad_norm": 0.2952023148536682,
|
|
"learning_rate": 7.006746065812895e-06,
|
|
"loss": 0.371,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"epoch": 6.5097236438075745,
|
|
"grad_norm": 0.3054288625717163,
|
|
"learning_rate": 6.973801770964031e-06,
|
|
"loss": 0.3666,
|
|
"step": 1055
|
|
},
|
|
{
|
|
"epoch": 6.54042988741044,
|
|
"grad_norm": 0.3081178069114685,
|
|
"learning_rate": 6.940755653591859e-06,
|
|
"loss": 0.3725,
|
|
"step": 1060
|
|
},
|
|
{
|
|
"epoch": 6.571136131013306,
|
|
"grad_norm": 0.30962446331977844,
|
|
"learning_rate": 6.907609418446623e-06,
|
|
"loss": 0.3678,
|
|
"step": 1065
|
|
},
|
|
{
|
|
"epoch": 6.601842374616172,
|
|
"grad_norm": 0.37370339035987854,
|
|
"learning_rate": 6.8743647754433485e-06,
|
|
"loss": 0.3718,
|
|
"step": 1070
|
|
},
|
|
{
|
|
"epoch": 6.632548618219038,
|
|
"grad_norm": 0.34246236085891724,
|
|
"learning_rate": 6.841023439573623e-06,
|
|
"loss": 0.3707,
|
|
"step": 1075
|
|
},
|
|
{
|
|
"epoch": 6.663254861821904,
|
|
"grad_norm": 0.3560137152671814,
|
|
"learning_rate": 6.807587130817134e-06,
|
|
"loss": 0.3702,
|
|
"step": 1080
|
|
},
|
|
{
|
|
"epoch": 6.69396110542477,
|
|
"grad_norm": 0.31180328130722046,
|
|
"learning_rate": 6.774057574052932e-06,
|
|
"loss": 0.3694,
|
|
"step": 1085
|
|
},
|
|
{
|
|
"epoch": 6.724667349027635,
|
|
"grad_norm": 0.3135690987110138,
|
|
"learning_rate": 6.740436498970453e-06,
|
|
"loss": 0.3685,
|
|
"step": 1090
|
|
},
|
|
{
|
|
"epoch": 6.755373592630502,
|
|
"grad_norm": 0.30377304553985596,
|
|
"learning_rate": 6.706725639980294e-06,
|
|
"loss": 0.3695,
|
|
"step": 1095
|
|
},
|
|
{
|
|
"epoch": 6.786079836233368,
|
|
"grad_norm": 0.30792489647865295,
|
|
"learning_rate": 6.6729267361247295e-06,
|
|
"loss": 0.3716,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 6.816786079836233,
|
|
"grad_norm": 0.3061661422252655,
|
|
"learning_rate": 6.639041530988009e-06,
|
|
"loss": 0.3694,
|
|
"step": 1105
|
|
},
|
|
{
|
|
"epoch": 6.847492323439099,
|
|
"grad_norm": 0.3261125683784485,
|
|
"learning_rate": 6.605071772606404e-06,
|
|
"loss": 0.3696,
|
|
"step": 1110
|
|
},
|
|
{
|
|
"epoch": 6.878198567041965,
|
|
"grad_norm": 0.295218288898468,
|
|
"learning_rate": 6.571019213378034e-06,
|
|
"loss": 0.3683,
|
|
"step": 1115
|
|
},
|
|
{
|
|
"epoch": 6.908904810644831,
|
|
"grad_norm": 0.30800992250442505,
|
|
"learning_rate": 6.536885609972467e-06,
|
|
"loss": 0.3711,
|
|
"step": 1120
|
|
},
|
|
{
|
|
"epoch": 6.939611054247697,
|
|
"grad_norm": 0.2889186143875122,
|
|
"learning_rate": 6.502672723240103e-06,
|
|
"loss": 0.3733,
|
|
"step": 1125
|
|
},
|
|
{
|
|
"epoch": 6.970317297850563,
|
|
"grad_norm": 0.30022063851356506,
|
|
"learning_rate": 6.4683823181213224e-06,
|
|
"loss": 0.3683,
|
|
"step": 1130
|
|
},
|
|
{
|
|
"epoch": 7.006141248720573,
|
|
"grad_norm": 0.7506215572357178,
|
|
"learning_rate": 6.434016163555452e-06,
|
|
"loss": 0.4364,
|
|
"step": 1135
|
|
},
|
|
{
|
|
"epoch": 7.036847492323439,
|
|
"grad_norm": 0.3902629315853119,
|
|
"learning_rate": 6.399576032389505e-06,
|
|
"loss": 0.3502,
|
|
"step": 1140
|
|
},
|
|
{
|
|
"epoch": 7.067553735926305,
|
|
"grad_norm": 0.3617767095565796,
|
|
"learning_rate": 6.365063701286728e-06,
|
|
"loss": 0.3541,
|
|
"step": 1145
|
|
},
|
|
{
|
|
"epoch": 7.098259979529171,
|
|
"grad_norm": 0.346937358379364,
|
|
"learning_rate": 6.330480950634942e-06,
|
|
"loss": 0.3529,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"epoch": 7.1289662231320365,
|
|
"grad_norm": 0.4673236310482025,
|
|
"learning_rate": 6.2958295644547026e-06,
|
|
"loss": 0.3524,
|
|
"step": 1155
|
|
},
|
|
{
|
|
"epoch": 7.159672466734903,
|
|
"grad_norm": 0.32648974657058716,
|
|
"learning_rate": 6.261111330307272e-06,
|
|
"loss": 0.3508,
|
|
"step": 1160
|
|
},
|
|
{
|
|
"epoch": 7.190378710337769,
|
|
"grad_norm": 0.31994733214378357,
|
|
"learning_rate": 6.22632803920239e-06,
|
|
"loss": 0.3491,
|
|
"step": 1165
|
|
},
|
|
{
|
|
"epoch": 7.221084953940634,
|
|
"grad_norm": 0.33214271068573,
|
|
"learning_rate": 6.191481485505898e-06,
|
|
"loss": 0.3491,
|
|
"step": 1170
|
|
},
|
|
{
|
|
"epoch": 7.2517911975435005,
|
|
"grad_norm": 0.33862602710723877,
|
|
"learning_rate": 6.1565734668471614e-06,
|
|
"loss": 0.3518,
|
|
"step": 1175
|
|
},
|
|
{
|
|
"epoch": 7.282497441146367,
|
|
"grad_norm": 0.3418470025062561,
|
|
"learning_rate": 6.121605784026339e-06,
|
|
"loss": 0.3552,
|
|
"step": 1180
|
|
},
|
|
{
|
|
"epoch": 7.313203684749232,
|
|
"grad_norm": 0.30670255422592163,
|
|
"learning_rate": 6.086580240921486e-06,
|
|
"loss": 0.3499,
|
|
"step": 1185
|
|
},
|
|
{
|
|
"epoch": 7.343909928352098,
|
|
"grad_norm": 0.2968249022960663,
|
|
"learning_rate": 6.051498644395496e-06,
|
|
"loss": 0.353,
|
|
"step": 1190
|
|
},
|
|
{
|
|
"epoch": 7.3746161719549645,
|
|
"grad_norm": 0.32068145275115967,
|
|
"learning_rate": 6.01636280420289e-06,
|
|
"loss": 0.3531,
|
|
"step": 1195
|
|
},
|
|
{
|
|
"epoch": 7.40532241555783,
|
|
"grad_norm": 0.30007636547088623,
|
|
"learning_rate": 5.981174532896459e-06,
|
|
"loss": 0.3505,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 7.436028659160696,
|
|
"grad_norm": 0.32065504789352417,
|
|
"learning_rate": 5.9459356457337556e-06,
|
|
"loss": 0.3488,
|
|
"step": 1205
|
|
},
|
|
{
|
|
"epoch": 7.466734902763562,
|
|
"grad_norm": 0.32740339636802673,
|
|
"learning_rate": 5.910647960583458e-06,
|
|
"loss": 0.3533,
|
|
"step": 1210
|
|
},
|
|
{
|
|
"epoch": 7.497441146366428,
|
|
"grad_norm": 0.3315354883670807,
|
|
"learning_rate": 5.875313297831579e-06,
|
|
"loss": 0.3538,
|
|
"step": 1215
|
|
},
|
|
{
|
|
"epoch": 7.528147389969294,
|
|
"grad_norm": 0.3044270873069763,
|
|
"learning_rate": 5.839933480287572e-06,
|
|
"loss": 0.3525,
|
|
"step": 1220
|
|
},
|
|
{
|
|
"epoch": 7.558853633572159,
|
|
"grad_norm": 0.3230973184108734,
|
|
"learning_rate": 5.804510333090287e-06,
|
|
"loss": 0.3525,
|
|
"step": 1225
|
|
},
|
|
{
|
|
"epoch": 7.5895598771750254,
|
|
"grad_norm": 0.305621862411499,
|
|
"learning_rate": 5.769045683613822e-06,
|
|
"loss": 0.3536,
|
|
"step": 1230
|
|
},
|
|
{
|
|
"epoch": 7.620266120777892,
|
|
"grad_norm": 0.30117788910865784,
|
|
"learning_rate": 5.733541361373253e-06,
|
|
"loss": 0.3531,
|
|
"step": 1235
|
|
},
|
|
{
|
|
"epoch": 7.650972364380758,
|
|
"grad_norm": 0.3083040118217468,
|
|
"learning_rate": 5.697999197930259e-06,
|
|
"loss": 0.3547,
|
|
"step": 1240
|
|
},
|
|
{
|
|
"epoch": 7.681678607983623,
|
|
"grad_norm": 0.3200486898422241,
|
|
"learning_rate": 5.662421026798624e-06,
|
|
"loss": 0.3541,
|
|
"step": 1245
|
|
},
|
|
{
|
|
"epoch": 7.7123848515864895,
|
|
"grad_norm": 0.3322235941886902,
|
|
"learning_rate": 5.626808683349672e-06,
|
|
"loss": 0.3539,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"epoch": 7.743091095189355,
|
|
"grad_norm": 0.3168823719024658,
|
|
"learning_rate": 5.591164004717567e-06,
|
|
"loss": 0.3571,
|
|
"step": 1255
|
|
},
|
|
{
|
|
"epoch": 7.773797338792221,
|
|
"grad_norm": 0.3052806854248047,
|
|
"learning_rate": 5.55548882970455e-06,
|
|
"loss": 0.3523,
|
|
"step": 1260
|
|
},
|
|
{
|
|
"epoch": 7.804503582395087,
|
|
"grad_norm": 0.3077663481235504,
|
|
"learning_rate": 5.519784998686081e-06,
|
|
"loss": 0.3553,
|
|
"step": 1265
|
|
},
|
|
{
|
|
"epoch": 7.835209825997953,
|
|
"grad_norm": 0.33488982915878296,
|
|
"learning_rate": 5.484054353515896e-06,
|
|
"loss": 0.3552,
|
|
"step": 1270
|
|
},
|
|
{
|
|
"epoch": 7.865916069600819,
|
|
"grad_norm": 0.31428563594818115,
|
|
"learning_rate": 5.448298737430992e-06,
|
|
"loss": 0.3539,
|
|
"step": 1275
|
|
},
|
|
{
|
|
"epoch": 7.896622313203685,
|
|
"grad_norm": 0.6753739714622498,
|
|
"learning_rate": 5.412519994956543e-06,
|
|
"loss": 0.3559,
|
|
"step": 1280
|
|
},
|
|
{
|
|
"epoch": 7.92732855680655,
|
|
"grad_norm": 0.31677567958831787,
|
|
"learning_rate": 5.376719971810741e-06,
|
|
"loss": 0.3546,
|
|
"step": 1285
|
|
},
|
|
{
|
|
"epoch": 7.958034800409417,
|
|
"grad_norm": 0.3080218732357025,
|
|
"learning_rate": 5.340900514809587e-06,
|
|
"loss": 0.3544,
|
|
"step": 1290
|
|
},
|
|
{
|
|
"epoch": 7.988741044012283,
|
|
"grad_norm": 0.31371742486953735,
|
|
"learning_rate": 5.305063471771614e-06,
|
|
"loss": 0.3548,
|
|
"step": 1295
|
|
},
|
|
{
|
|
"epoch": 8.024564994882292,
|
|
"grad_norm": 0.40025079250335693,
|
|
"learning_rate": 5.26921069142257e-06,
|
|
"loss": 0.4121,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 8.055271238485158,
|
|
"grad_norm": 0.35833466053009033,
|
|
"learning_rate": 5.233344023300037e-06,
|
|
"loss": 0.3357,
|
|
"step": 1305
|
|
},
|
|
{
|
|
"epoch": 8.085977482088024,
|
|
"grad_norm": 0.3553675711154938,
|
|
"learning_rate": 5.197465317658036e-06,
|
|
"loss": 0.3353,
|
|
"step": 1310
|
|
},
|
|
{
|
|
"epoch": 8.11668372569089,
|
|
"grad_norm": 0.3430934548377991,
|
|
"learning_rate": 5.161576425371554e-06,
|
|
"loss": 0.3347,
|
|
"step": 1315
|
|
},
|
|
{
|
|
"epoch": 8.147389969293757,
|
|
"grad_norm": 0.33370262384414673,
|
|
"learning_rate": 5.125679197841088e-06,
|
|
"loss": 0.3362,
|
|
"step": 1320
|
|
},
|
|
{
|
|
"epoch": 8.178096212896623,
|
|
"grad_norm": 0.32513174414634705,
|
|
"learning_rate": 5.089775486897121e-06,
|
|
"loss": 0.3368,
|
|
"step": 1325
|
|
},
|
|
{
|
|
"epoch": 8.208802456499487,
|
|
"grad_norm": 0.35142919421195984,
|
|
"learning_rate": 5.053867144704594e-06,
|
|
"loss": 0.3359,
|
|
"step": 1330
|
|
},
|
|
{
|
|
"epoch": 8.239508700102354,
|
|
"grad_norm": 0.32932382822036743,
|
|
"learning_rate": 5.017956023667363e-06,
|
|
"loss": 0.3359,
|
|
"step": 1335
|
|
},
|
|
{
|
|
"epoch": 8.27021494370522,
|
|
"grad_norm": 0.32409489154815674,
|
|
"learning_rate": 4.982043976332638e-06,
|
|
"loss": 0.3365,
|
|
"step": 1340
|
|
},
|
|
{
|
|
"epoch": 8.300921187308086,
|
|
"grad_norm": 0.32708844542503357,
|
|
"learning_rate": 4.946132855295407e-06,
|
|
"loss": 0.3389,
|
|
"step": 1345
|
|
},
|
|
{
|
|
"epoch": 8.331627430910952,
|
|
"grad_norm": 0.30396559834480286,
|
|
"learning_rate": 4.910224513102881e-06,
|
|
"loss": 0.3361,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"epoch": 8.362333674513819,
|
|
"grad_norm": 0.3169805109500885,
|
|
"learning_rate": 4.8743208021589135e-06,
|
|
"loss": 0.338,
|
|
"step": 1355
|
|
},
|
|
{
|
|
"epoch": 8.393039918116683,
|
|
"grad_norm": 0.32971227169036865,
|
|
"learning_rate": 4.838423574628447e-06,
|
|
"loss": 0.3371,
|
|
"step": 1360
|
|
},
|
|
{
|
|
"epoch": 8.42374616171955,
|
|
"grad_norm": 0.43314287066459656,
|
|
"learning_rate": 4.802534682341966e-06,
|
|
"loss": 0.3376,
|
|
"step": 1365
|
|
},
|
|
{
|
|
"epoch": 8.454452405322415,
|
|
"grad_norm": 0.3369028866291046,
|
|
"learning_rate": 4.7666559766999635e-06,
|
|
"loss": 0.3376,
|
|
"step": 1370
|
|
},
|
|
{
|
|
"epoch": 8.485158648925282,
|
|
"grad_norm": 0.3581888973712921,
|
|
"learning_rate": 4.730789308577432e-06,
|
|
"loss": 0.3391,
|
|
"step": 1375
|
|
},
|
|
{
|
|
"epoch": 8.515864892528148,
|
|
"grad_norm": 0.4204093813896179,
|
|
"learning_rate": 4.694936528228387e-06,
|
|
"loss": 0.3378,
|
|
"step": 1380
|
|
},
|
|
{
|
|
"epoch": 8.546571136131014,
|
|
"grad_norm": 0.3086753487586975,
|
|
"learning_rate": 4.659099485190414e-06,
|
|
"loss": 0.336,
|
|
"step": 1385
|
|
},
|
|
{
|
|
"epoch": 8.577277379733879,
|
|
"grad_norm": 0.3337133526802063,
|
|
"learning_rate": 4.6232800281892604e-06,
|
|
"loss": 0.3379,
|
|
"step": 1390
|
|
},
|
|
{
|
|
"epoch": 8.607983623336745,
|
|
"grad_norm": 0.313374787569046,
|
|
"learning_rate": 4.587480005043458e-06,
|
|
"loss": 0.3385,
|
|
"step": 1395
|
|
},
|
|
{
|
|
"epoch": 8.638689866939611,
|
|
"grad_norm": 0.3218012750148773,
|
|
"learning_rate": 4.551701262569009e-06,
|
|
"loss": 0.336,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 8.669396110542477,
|
|
"grad_norm": 0.3326752781867981,
|
|
"learning_rate": 4.515945646484105e-06,
|
|
"loss": 0.3407,
|
|
"step": 1405
|
|
},
|
|
{
|
|
"epoch": 8.700102354145343,
|
|
"grad_norm": 0.31090524792671204,
|
|
"learning_rate": 4.480215001313919e-06,
|
|
"loss": 0.3383,
|
|
"step": 1410
|
|
},
|
|
{
|
|
"epoch": 8.730808597748208,
|
|
"grad_norm": 0.3121759593486786,
|
|
"learning_rate": 4.444511170295451e-06,
|
|
"loss": 0.3394,
|
|
"step": 1415
|
|
},
|
|
{
|
|
"epoch": 8.761514841351074,
|
|
"grad_norm": 0.3316711485385895,
|
|
"learning_rate": 4.408835995282434e-06,
|
|
"loss": 0.3406,
|
|
"step": 1420
|
|
},
|
|
{
|
|
"epoch": 8.79222108495394,
|
|
"grad_norm": 0.3368071913719177,
|
|
"learning_rate": 4.373191316650328e-06,
|
|
"loss": 0.3392,
|
|
"step": 1425
|
|
},
|
|
{
|
|
"epoch": 8.822927328556807,
|
|
"grad_norm": 0.3207019567489624,
|
|
"learning_rate": 4.3375789732013775e-06,
|
|
"loss": 0.3395,
|
|
"step": 1430
|
|
},
|
|
{
|
|
"epoch": 8.853633572159673,
|
|
"grad_norm": 0.3298065662384033,
|
|
"learning_rate": 4.302000802069744e-06,
|
|
"loss": 0.3413,
|
|
"step": 1435
|
|
},
|
|
{
|
|
"epoch": 8.884339815762539,
|
|
"grad_norm": 0.30852875113487244,
|
|
"learning_rate": 4.2664586386267474e-06,
|
|
"loss": 0.3417,
|
|
"step": 1440
|
|
},
|
|
{
|
|
"epoch": 8.915046059365404,
|
|
"grad_norm": 0.31210556626319885,
|
|
"learning_rate": 4.230954316386179e-06,
|
|
"loss": 0.336,
|
|
"step": 1445
|
|
},
|
|
{
|
|
"epoch": 8.94575230296827,
|
|
"grad_norm": 0.3171374797821045,
|
|
"learning_rate": 4.195489666909714e-06,
|
|
"loss": 0.3428,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"epoch": 8.976458546571136,
|
|
"grad_norm": 0.32448017597198486,
|
|
"learning_rate": 4.160066519712428e-06,
|
|
"loss": 0.3413,
|
|
"step": 1455
|
|
},
|
|
{
|
|
"epoch": 9.012282497441147,
|
|
"grad_norm": 0.39366790652275085,
|
|
"learning_rate": 4.1246867021684206e-06,
|
|
"loss": 0.401,
|
|
"step": 1460
|
|
},
|
|
{
|
|
"epoch": 9.042988741044013,
|
|
"grad_norm": 0.4089623987674713,
|
|
"learning_rate": 4.089352039416543e-06,
|
|
"loss": 0.3221,
|
|
"step": 1465
|
|
},
|
|
{
|
|
"epoch": 9.073694984646878,
|
|
"grad_norm": 0.5040475130081177,
|
|
"learning_rate": 4.054064354266244e-06,
|
|
"loss": 0.3216,
|
|
"step": 1470
|
|
},
|
|
{
|
|
"epoch": 9.104401228249744,
|
|
"grad_norm": 0.3531099557876587,
|
|
"learning_rate": 4.018825467103542e-06,
|
|
"loss": 0.3208,
|
|
"step": 1475
|
|
},
|
|
{
|
|
"epoch": 9.13510747185261,
|
|
"grad_norm": 0.34139832854270935,
|
|
"learning_rate": 3.983637195797111e-06,
|
|
"loss": 0.3202,
|
|
"step": 1480
|
|
},
|
|
{
|
|
"epoch": 9.165813715455476,
|
|
"grad_norm": 0.3356687128543854,
|
|
"learning_rate": 3.948501355604507e-06,
|
|
"loss": 0.3236,
|
|
"step": 1485
|
|
},
|
|
{
|
|
"epoch": 9.196519959058342,
|
|
"grad_norm": 0.34154486656188965,
|
|
"learning_rate": 3.9134197590785164e-06,
|
|
"loss": 0.3201,
|
|
"step": 1490
|
|
},
|
|
{
|
|
"epoch": 9.227226202661209,
|
|
"grad_norm": 0.32101234793663025,
|
|
"learning_rate": 3.878394215973663e-06,
|
|
"loss": 0.3229,
|
|
"step": 1495
|
|
},
|
|
{
|
|
"epoch": 9.257932446264073,
|
|
"grad_norm": 0.31901758909225464,
|
|
"learning_rate": 3.843426533152841e-06,
|
|
"loss": 0.3235,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 9.28863868986694,
|
|
"grad_norm": 0.4290211796760559,
|
|
"learning_rate": 3.808518514494105e-06,
|
|
"loss": 0.3209,
|
|
"step": 1505
|
|
},
|
|
{
|
|
"epoch": 9.319344933469806,
|
|
"grad_norm": 0.34095343947410583,
|
|
"learning_rate": 3.773671960797613e-06,
|
|
"loss": 0.323,
|
|
"step": 1510
|
|
},
|
|
{
|
|
"epoch": 9.350051177072672,
|
|
"grad_norm": 0.4352162182331085,
|
|
"learning_rate": 3.7388886696927317e-06,
|
|
"loss": 0.321,
|
|
"step": 1515
|
|
},
|
|
{
|
|
"epoch": 9.380757420675538,
|
|
"grad_norm": 0.3285370171070099,
|
|
"learning_rate": 3.704170435545299e-06,
|
|
"loss": 0.3227,
|
|
"step": 1520
|
|
},
|
|
{
|
|
"epoch": 9.411463664278402,
|
|
"grad_norm": 0.31492966413497925,
|
|
"learning_rate": 3.6695190493650608e-06,
|
|
"loss": 0.3238,
|
|
"step": 1525
|
|
},
|
|
{
|
|
"epoch": 9.442169907881269,
|
|
"grad_norm": 0.3235698938369751,
|
|
"learning_rate": 3.634936298713274e-06,
|
|
"loss": 0.3249,
|
|
"step": 1530
|
|
},
|
|
{
|
|
"epoch": 9.472876151484135,
|
|
"grad_norm": 0.3167796730995178,
|
|
"learning_rate": 3.6004239676104957e-06,
|
|
"loss": 0.3252,
|
|
"step": 1535
|
|
},
|
|
{
|
|
"epoch": 9.503582395087001,
|
|
"grad_norm": 0.3134520649909973,
|
|
"learning_rate": 3.5659838364445505e-06,
|
|
"loss": 0.3233,
|
|
"step": 1540
|
|
},
|
|
{
|
|
"epoch": 9.534288638689867,
|
|
"grad_norm": 0.3369145691394806,
|
|
"learning_rate": 3.5316176818786797e-06,
|
|
"loss": 0.3253,
|
|
"step": 1545
|
|
},
|
|
{
|
|
"epoch": 9.564994882292734,
|
|
"grad_norm": 0.313290536403656,
|
|
"learning_rate": 3.497327276759899e-06,
|
|
"loss": 0.3275,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"epoch": 9.595701125895598,
|
|
"grad_norm": 0.3463856279850006,
|
|
"learning_rate": 3.463114390027533e-06,
|
|
"loss": 0.3247,
|
|
"step": 1555
|
|
},
|
|
{
|
|
"epoch": 9.626407369498464,
|
|
"grad_norm": 0.3312392234802246,
|
|
"learning_rate": 3.4289807866219683e-06,
|
|
"loss": 0.3271,
|
|
"step": 1560
|
|
},
|
|
{
|
|
"epoch": 9.65711361310133,
|
|
"grad_norm": 0.32030391693115234,
|
|
"learning_rate": 3.394928227393598e-06,
|
|
"loss": 0.3253,
|
|
"step": 1565
|
|
},
|
|
{
|
|
"epoch": 9.687819856704197,
|
|
"grad_norm": 0.3277048170566559,
|
|
"learning_rate": 3.3609584690119924e-06,
|
|
"loss": 0.325,
|
|
"step": 1570
|
|
},
|
|
{
|
|
"epoch": 9.718526100307063,
|
|
"grad_norm": 0.3145771920681,
|
|
"learning_rate": 3.3270732638752713e-06,
|
|
"loss": 0.3238,
|
|
"step": 1575
|
|
},
|
|
{
|
|
"epoch": 9.749232343909929,
|
|
"grad_norm": 0.33325299620628357,
|
|
"learning_rate": 3.293274360019707e-06,
|
|
"loss": 0.3262,
|
|
"step": 1580
|
|
},
|
|
{
|
|
"epoch": 9.779938587512794,
|
|
"grad_norm": 0.3249775469303131,
|
|
"learning_rate": 3.259563501029548e-06,
|
|
"loss": 0.3263,
|
|
"step": 1585
|
|
},
|
|
{
|
|
"epoch": 9.81064483111566,
|
|
"grad_norm": 0.3150160610675812,
|
|
"learning_rate": 3.2259424259470705e-06,
|
|
"loss": 0.3248,
|
|
"step": 1590
|
|
},
|
|
{
|
|
"epoch": 9.841351074718526,
|
|
"grad_norm": 0.3632504343986511,
|
|
"learning_rate": 3.1924128691828678e-06,
|
|
"loss": 0.3234,
|
|
"step": 1595
|
|
},
|
|
{
|
|
"epoch": 9.872057318321392,
|
|
"grad_norm": 0.32622626423835754,
|
|
"learning_rate": 3.158976560426379e-06,
|
|
"loss": 0.324,
|
|
"step": 1600
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 2430,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 15,
|
|
"save_steps": 400,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": false
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 3.48962200614737e+19,
|
|
"train_batch_size": 1,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|