9b-43 / trainer_state.json
furproxy's picture
Upload folder using huggingface_hub
95141f8 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 690,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.008695652173913044,
"grad_norm": 0.732024610042572,
"learning_rate": 8.571428571428571e-07,
"loss": 2.46012020111084,
"step": 2
},
{
"epoch": 0.017391304347826087,
"grad_norm": 1.1184877157211304,
"learning_rate": 2.5714285714285716e-06,
"loss": 2.301504135131836,
"step": 4
},
{
"epoch": 0.02608695652173913,
"grad_norm": 3.1666553020477295,
"learning_rate": 4.2857142857142855e-06,
"loss": 2.9570844173431396,
"step": 6
},
{
"epoch": 0.034782608695652174,
"grad_norm": 0.25166502594947815,
"learning_rate": 6e-06,
"loss": 1.8961513042449951,
"step": 8
},
{
"epoch": 0.043478260869565216,
"grad_norm": 0.20189805328845978,
"learning_rate": 7.714285714285714e-06,
"loss": 3.0871455669403076,
"step": 10
},
{
"epoch": 0.05217391304347826,
"grad_norm": 0.20353659987449646,
"learning_rate": 9.428571428571428e-06,
"loss": 2.292584180831909,
"step": 12
},
{
"epoch": 0.06086956521739131,
"grad_norm": 0.6789005994796753,
"learning_rate": 1.1142857142857143e-05,
"loss": 2.5501410961151123,
"step": 14
},
{
"epoch": 0.06956521739130435,
"grad_norm": 0.2317086160182953,
"learning_rate": 1.2857142857142857e-05,
"loss": 1.931159257888794,
"step": 16
},
{
"epoch": 0.0782608695652174,
"grad_norm": 1.7764248847961426,
"learning_rate": 1.4571428571428571e-05,
"loss": 2.571274518966675,
"step": 18
},
{
"epoch": 0.08695652173913043,
"grad_norm": 0.4352605640888214,
"learning_rate": 1.6285714285714283e-05,
"loss": 1.9101747274398804,
"step": 20
},
{
"epoch": 0.09565217391304348,
"grad_norm": 0.18403619527816772,
"learning_rate": 1.8e-05,
"loss": 1.8424170017242432,
"step": 22
},
{
"epoch": 0.10434782608695652,
"grad_norm": 0.16127634048461914,
"learning_rate": 1.9714285714285714e-05,
"loss": 5.551755905151367,
"step": 24
},
{
"epoch": 0.11304347826086956,
"grad_norm": 0.1956806778907776,
"learning_rate": 2.1428571428571428e-05,
"loss": 1.828694224357605,
"step": 26
},
{
"epoch": 0.12173913043478261,
"grad_norm": 2.0333733558654785,
"learning_rate": 2.3142857142857145e-05,
"loss": 2.4214303493499756,
"step": 28
},
{
"epoch": 0.13043478260869565,
"grad_norm": 0.24648341536521912,
"learning_rate": 2.485714285714286e-05,
"loss": 3.56320858001709,
"step": 30
},
{
"epoch": 0.1391304347826087,
"grad_norm": 0.17006301879882812,
"learning_rate": 2.657142857142857e-05,
"loss": 1.7686303853988647,
"step": 32
},
{
"epoch": 0.14782608695652175,
"grad_norm": 1.2820721864700317,
"learning_rate": 2.8285714285714287e-05,
"loss": 1.9711768627166748,
"step": 34
},
{
"epoch": 0.1565217391304348,
"grad_norm": 1.2253106832504272,
"learning_rate": 3e-05,
"loss": 1.890865683555603,
"step": 36
},
{
"epoch": 0.16521739130434782,
"grad_norm": 0.15150298178195953,
"learning_rate": 2.9994720778760907e-05,
"loss": 1.682741641998291,
"step": 38
},
{
"epoch": 0.17391304347826086,
"grad_norm": 0.14206662774085999,
"learning_rate": 2.9978887486836082e-05,
"loss": 1.8588255643844604,
"step": 40
},
{
"epoch": 0.1826086956521739,
"grad_norm": 0.18157699704170227,
"learning_rate": 2.9952513235982553e-05,
"loss": 1.7579265832901,
"step": 42
},
{
"epoch": 0.19130434782608696,
"grad_norm": 0.13783733546733856,
"learning_rate": 2.991561986706388e-05,
"loss": 3.504922389984131,
"step": 44
},
{
"epoch": 0.2,
"grad_norm": 0.4945126473903656,
"learning_rate": 2.9868237931963492e-05,
"loss": 1.5656976699829102,
"step": 46
},
{
"epoch": 0.20869565217391303,
"grad_norm": 1.3513635396957397,
"learning_rate": 2.9810406668284216e-05,
"loss": 1.2987524271011353,
"step": 48
},
{
"epoch": 0.21739130434782608,
"grad_norm": 0.3287999629974365,
"learning_rate": 2.9742173966855167e-05,
"loss": 1.316739797592163,
"step": 50
},
{
"epoch": 0.22608695652173913,
"grad_norm": 0.6960898041725159,
"learning_rate": 2.9663596332072664e-05,
"loss": 1.020365595817566,
"step": 52
},
{
"epoch": 0.23478260869565218,
"grad_norm": 0.3621000051498413,
"learning_rate": 2.9574738835108235e-05,
"loss": 0.9612956047058105,
"step": 54
},
{
"epoch": 0.24347826086956523,
"grad_norm": 0.36931225657463074,
"learning_rate": 2.94756750600223e-05,
"loss": 1.4254815578460693,
"step": 56
},
{
"epoch": 0.25217391304347825,
"grad_norm": 0.11330584436655045,
"learning_rate": 2.9366487042828275e-05,
"loss": 1.4714879989624023,
"step": 58
},
{
"epoch": 0.2608695652173913,
"grad_norm": 0.13919749855995178,
"learning_rate": 2.9247265203557428e-05,
"loss": 1.8500466346740723,
"step": 60
},
{
"epoch": 0.26956521739130435,
"grad_norm": 1.1177152395248413,
"learning_rate": 2.9118108271380923e-05,
"loss": 1.0420849323272705,
"step": 62
},
{
"epoch": 0.2782608695652174,
"grad_norm": 0.14357954263687134,
"learning_rate": 2.8979123202850823e-05,
"loss": 1.2155001163482666,
"step": 64
},
{
"epoch": 0.28695652173913044,
"grad_norm": 0.29200610518455505,
"learning_rate": 2.8830425093328024e-05,
"loss": 0.5274585485458374,
"step": 66
},
{
"epoch": 0.2956521739130435,
"grad_norm": 0.14065444469451904,
"learning_rate": 2.8672137081670264e-05,
"loss": 0.9751191735267639,
"step": 68
},
{
"epoch": 0.30434782608695654,
"grad_norm": 1.254546046257019,
"learning_rate": 2.850439024825919e-05,
"loss": 0.45779138803482056,
"step": 70
},
{
"epoch": 0.3130434782608696,
"grad_norm": 2.3761823177337646,
"learning_rate": 2.8327323506450993e-05,
"loss": 1.0098928213119507,
"step": 72
},
{
"epoch": 0.3217391304347826,
"grad_norm": 0.19490118324756622,
"learning_rate": 2.8141083487540432e-05,
"loss": 1.4895820617675781,
"step": 74
},
{
"epoch": 0.33043478260869563,
"grad_norm": 1.0330501794815063,
"learning_rate": 2.794582441933354e-05,
"loss": 0.339144766330719,
"step": 76
},
{
"epoch": 0.3391304347826087,
"grad_norm": 0.17919474840164185,
"learning_rate": 2.7741707998429524e-05,
"loss": 1.4342169761657715,
"step": 78
},
{
"epoch": 0.34782608695652173,
"grad_norm": 0.189740389585495,
"learning_rate": 2.7528903256317714e-05,
"loss": 1.676353931427002,
"step": 80
},
{
"epoch": 0.3565217391304348,
"grad_norm": 0.152068629860878,
"learning_rate": 2.7307586419400365e-05,
"loss": 1.4120248556137085,
"step": 82
},
{
"epoch": 0.3652173913043478,
"grad_norm": 0.11617185920476913,
"learning_rate": 2.707794076305724e-05,
"loss": 1.4109995365142822,
"step": 84
},
{
"epoch": 0.3739130434782609,
"grad_norm": 0.5076650977134705,
"learning_rate": 2.6840156459872887e-05,
"loss": 1.8156630992889404,
"step": 86
},
{
"epoch": 0.3826086956521739,
"grad_norm": 0.3797202408313751,
"learning_rate": 2.6594430422152184e-05,
"loss": 1.2095890045166016,
"step": 88
},
{
"epoch": 0.391304347826087,
"grad_norm": 0.1064012348651886,
"learning_rate": 2.6340966138854725e-05,
"loss": 1.3283032178878784,
"step": 90
},
{
"epoch": 0.4,
"grad_norm": 0.2617345452308655,
"learning_rate": 2.607997350708289e-05,
"loss": 0.3713761866092682,
"step": 92
},
{
"epoch": 0.40869565217391307,
"grad_norm": 0.14992359280586243,
"learning_rate": 2.5811668658263345e-05,
"loss": 1.2495781183242798,
"step": 94
},
{
"epoch": 0.41739130434782606,
"grad_norm": 0.8707973957061768,
"learning_rate": 2.5536273779165707e-05,
"loss": 0.892501950263977,
"step": 96
},
{
"epoch": 0.4260869565217391,
"grad_norm": 0.18348021805286407,
"learning_rate": 2.5254016927906833e-05,
"loss": 0.99811190366745,
"step": 98
},
{
"epoch": 0.43478260869565216,
"grad_norm": 0.28490695357322693,
"learning_rate": 2.4965131845092846e-05,
"loss": 0.9439688324928284,
"step": 100
},
{
"epoch": 0.4434782608695652,
"grad_norm": 0.19018062949180603,
"learning_rate": 2.466985776025548e-05,
"loss": 1.434984803199768,
"step": 102
},
{
"epoch": 0.45217391304347826,
"grad_norm": 0.3262655436992645,
"learning_rate": 2.4368439193742983e-05,
"loss": 1.6902875900268555,
"step": 104
},
{
"epoch": 0.4608695652173913,
"grad_norm": 0.1134955883026123,
"learning_rate": 2.4061125754229576e-05,
"loss": 0.3920702636241913,
"step": 106
},
{
"epoch": 0.46956521739130436,
"grad_norm": 0.15943671762943268,
"learning_rate": 2.3748171932011248e-05,
"loss": 1.1952660083770752,
"step": 108
},
{
"epoch": 0.4782608695652174,
"grad_norm": 0.17980021238327026,
"learning_rate": 2.3429836888259e-05,
"loss": 1.0627365112304688,
"step": 110
},
{
"epoch": 0.48695652173913045,
"grad_norm": 0.20388175547122955,
"learning_rate": 2.310638424040406e-05,
"loss": 1.4079805612564087,
"step": 112
},
{
"epoch": 0.4956521739130435,
"grad_norm": 0.0773826614022255,
"learning_rate": 2.2778081843832856e-05,
"loss": 0.8712098002433777,
"step": 114
},
{
"epoch": 0.5043478260869565,
"grad_norm": 0.1742291897535324,
"learning_rate": 2.2445201570072414e-05,
"loss": 1.3538367748260498,
"step": 116
},
{
"epoch": 0.5130434782608696,
"grad_norm": 0.15458093583583832,
"learning_rate": 2.2108019081649993e-05,
"loss": 1.612799048423767,
"step": 118
},
{
"epoch": 0.5217391304347826,
"grad_norm": 0.5128660202026367,
"learning_rate": 2.1766813603813334e-05,
"loss": 0.8653756380081177,
"step": 120
},
{
"epoch": 0.5304347826086957,
"grad_norm": 0.6170827150344849,
"learning_rate": 2.1421867693300552e-05,
"loss": 0.7663818597793579,
"step": 122
},
{
"epoch": 0.5391304347826087,
"grad_norm": 0.17489704489707947,
"learning_rate": 2.1073467004351202e-05,
"loss": 0.7496420741081238,
"step": 124
},
{
"epoch": 0.5478260869565217,
"grad_norm": 0.7850053906440735,
"learning_rate": 2.0721900052152227e-05,
"loss": 0.7726188898086548,
"step": 126
},
{
"epoch": 0.5565217391304348,
"grad_norm": 0.09144008904695511,
"learning_rate": 2.036745797391477e-05,
"loss": 0.7957323789596558,
"step": 128
},
{
"epoch": 0.5652173913043478,
"grad_norm": 0.11362311989068985,
"learning_rate": 2.0010434287779607e-05,
"loss": 1.3306431770324707,
"step": 130
},
{
"epoch": 0.5739130434782609,
"grad_norm": 0.13353990018367767,
"learning_rate": 1.965112464975093e-05,
"loss": 1.3570564985275269,
"step": 132
},
{
"epoch": 0.5826086956521739,
"grad_norm": 0.16489805281162262,
"learning_rate": 1.928982660885971e-05,
"loss": 1.3512859344482422,
"step": 134
},
{
"epoch": 0.591304347826087,
"grad_norm": 0.12622074782848358,
"learning_rate": 1.8926839360759467e-05,
"loss": 1.3253053426742554,
"step": 136
},
{
"epoch": 0.6,
"grad_norm": 0.18052835762500763,
"learning_rate": 1.856246349995838e-05,
"loss": 1.363409161567688,
"step": 138
},
{
"epoch": 0.6086956521739131,
"grad_norm": 1.0225669145584106,
"learning_rate": 1.8197000770893056e-05,
"loss": 0.9811846613883972,
"step": 140
},
{
"epoch": 0.6173913043478261,
"grad_norm": 0.12899279594421387,
"learning_rate": 1.7830753818049987e-05,
"loss": 1.3022255897521973,
"step": 142
},
{
"epoch": 0.6260869565217392,
"grad_norm": 0.15879316627979279,
"learning_rate": 1.7464025935341674e-05,
"loss": 0.7987528443336487,
"step": 144
},
{
"epoch": 0.6347826086956522,
"grad_norm": 0.15264292061328888,
"learning_rate": 1.709712081494495e-05,
"loss": 0.790221095085144,
"step": 146
},
{
"epoch": 0.6434782608695652,
"grad_norm": 0.24214515089988708,
"learning_rate": 1.673034229580953e-05,
"loss": 0.8017470240592957,
"step": 148
},
{
"epoch": 0.6521739130434783,
"grad_norm": 0.6817665696144104,
"learning_rate": 1.6363994112044934e-05,
"loss": 0.8057210445404053,
"step": 150
},
{
"epoch": 0.6608695652173913,
"grad_norm": 0.2915583848953247,
"learning_rate": 1.5998379641394322e-05,
"loss": 0.8728779554367065,
"step": 152
},
{
"epoch": 0.6695652173913044,
"grad_norm": 0.22204452753067017,
"learning_rate": 1.5633801654003383e-05,
"loss": 1.3450850248336792,
"step": 154
},
{
"epoch": 0.6782608695652174,
"grad_norm": 0.13754141330718994,
"learning_rate": 1.5270562061692413e-05,
"loss": 1.295073390007019,
"step": 156
},
{
"epoch": 0.6869565217391305,
"grad_norm": 0.1301906257867813,
"learning_rate": 1.4908961667939148e-05,
"loss": 1.3460811376571655,
"step": 158
},
{
"epoch": 0.6956521739130435,
"grad_norm": 0.2238963395357132,
"learning_rate": 1.4549299918779476e-05,
"loss": 1.4722148180007935,
"step": 160
},
{
"epoch": 0.7043478260869566,
"grad_norm": 0.1082434207201004,
"learning_rate": 1.4191874654832224e-05,
"loss": 1.301584243774414,
"step": 162
},
{
"epoch": 0.7130434782608696,
"grad_norm": 0.29728466272354126,
"learning_rate": 1.3836981864653448e-05,
"loss": 0.8648190498352051,
"step": 164
},
{
"epoch": 0.7217391304347827,
"grad_norm": 0.12261836230754852,
"learning_rate": 1.3484915439624399e-05,
"loss": 1.2689242362976074,
"step": 166
},
{
"epoch": 0.7304347826086957,
"grad_norm": 0.09543546289205551,
"learning_rate": 1.3135966930576233e-05,
"loss": 0.9420109987258911,
"step": 168
},
{
"epoch": 0.7391304347826086,
"grad_norm": 0.1176115870475769,
"learning_rate": 1.2790425306352913e-05,
"loss": 1.3050771951675415,
"step": 170
},
{
"epoch": 0.7478260869565218,
"grad_norm": 0.16124227643013,
"learning_rate": 1.2448576714512348e-05,
"loss": 1.160625696182251,
"step": 172
},
{
"epoch": 0.7565217391304347,
"grad_norm": 0.45197612047195435,
"learning_rate": 1.2110704244363801e-05,
"loss": 0.5953898429870605,
"step": 174
},
{
"epoch": 0.7652173913043478,
"grad_norm": 0.09739059209823608,
"learning_rate": 1.1777087692537941e-05,
"loss": 0.6910260915756226,
"step": 176
},
{
"epoch": 0.7739130434782608,
"grad_norm": 0.1326562613248825,
"learning_rate": 1.1448003331283567e-05,
"loss": 1.332924723625183,
"step": 178
},
{
"epoch": 0.782608695652174,
"grad_norm": 0.3839710056781769,
"learning_rate": 1.1123723679682947e-05,
"loss": 0.37987953424453735,
"step": 180
},
{
"epoch": 0.7913043478260869,
"grad_norm": 0.20606131851673126,
"learning_rate": 1.0804517277975145e-05,
"loss": 0.6722633242607117,
"step": 182
},
{
"epoch": 0.8,
"grad_norm": 0.2714339792728424,
"learning_rate": 1.0490648465174374e-05,
"loss": 0.6781936883926392,
"step": 184
},
{
"epoch": 0.808695652173913,
"grad_norm": 0.12319911271333694,
"learning_rate": 1.0182377160167355e-05,
"loss": 1.3093880414962769,
"step": 186
},
{
"epoch": 0.8173913043478261,
"grad_norm": 0.19812533259391785,
"learning_rate": 9.87995864647107e-06,
"loss": 1.0597083568572998,
"step": 188
},
{
"epoch": 0.8260869565217391,
"grad_norm": 0.18480421602725983,
"learning_rate": 9.583643360829146e-06,
"loss": 1.113531231880188,
"step": 190
},
{
"epoch": 0.8347826086956521,
"grad_norm": 0.170735701918602,
"learning_rate": 9.293676685821879e-06,
"loss": 0.7899190783500671,
"step": 192
},
{
"epoch": 0.8434782608695652,
"grad_norm": 0.16350802779197693,
"learning_rate": 9.010298746661722e-06,
"loss": 1.299129605293274,
"step": 194
},
{
"epoch": 0.8521739130434782,
"grad_norm": 0.18414396047592163,
"learning_rate": 8.733744212342408e-06,
"loss": 1.3935397863388062,
"step": 196
},
{
"epoch": 0.8608695652173913,
"grad_norm": 0.2395719736814499,
"learning_rate": 8.46424210130651e-06,
"loss": 1.3050575256347656,
"step": 198
},
{
"epoch": 0.8695652173913043,
"grad_norm": 0.33917221426963806,
"learning_rate": 8.202015591792223e-06,
"loss": 0.6221288442611694,
"step": 200
},
{
"epoch": 0.8782608695652174,
"grad_norm": 0.24979734420776367,
"learning_rate": 7.947281837016573e-06,
"loss": 0.8170241117477417,
"step": 202
},
{
"epoch": 0.8869565217391304,
"grad_norm": 0.06852871179580688,
"learning_rate": 7.70025178534795e-06,
"loss": 0.729548990726471,
"step": 204
},
{
"epoch": 0.8956521739130435,
"grad_norm": 3.180227756500244,
"learning_rate": 7.461130005617023e-06,
"loss": 0.3102116882801056,
"step": 206
},
{
"epoch": 0.9043478260869565,
"grad_norm": 0.5475708246231079,
"learning_rate": 7.230114517710568e-06,
"loss": 1.066127061843872,
"step": 208
},
{
"epoch": 0.9130434782608695,
"grad_norm": 0.21181726455688477,
"learning_rate": 7.007396628588649e-06,
"loss": 1.1609755754470825,
"step": 210
},
{
"epoch": 0.9217391304347826,
"grad_norm": 0.3930780589580536,
"learning_rate": 6.793160773860759e-06,
"loss": 1.2955363988876343,
"step": 212
},
{
"epoch": 0.9304347826086956,
"grad_norm": 0.2640330195426941,
"learning_rate": 6.587584365052308e-06,
"loss": 0.22745762765407562,
"step": 214
},
{
"epoch": 0.9391304347826087,
"grad_norm": 0.5837603807449341,
"learning_rate": 6.390837642687785e-06,
"loss": 0.25208932161331177,
"step": 216
},
{
"epoch": 0.9478260869565217,
"grad_norm": 0.09962380677461624,
"learning_rate": 6.203083535312356e-06,
"loss": 0.7900155186653137,
"step": 218
},
{
"epoch": 0.9565217391304348,
"grad_norm": 0.6199167966842651,
"learning_rate": 6.024477524568591e-06,
"loss": 0.8592709302902222,
"step": 220
},
{
"epoch": 0.9652173913043478,
"grad_norm": 0.2959963381290436,
"learning_rate": 5.855167516440076e-06,
"loss": 1.3669737577438354,
"step": 222
},
{
"epoch": 0.9739130434782609,
"grad_norm": 0.13669747114181519,
"learning_rate": 5.695293718768518e-06,
"loss": 1.4519469738006592,
"step": 224
},
{
"epoch": 0.9826086956521739,
"grad_norm": 0.15183264017105103,
"learning_rate": 5.5449885251458e-06,
"loss": 1.3087224960327148,
"step": 226
},
{
"epoch": 0.991304347826087,
"grad_norm": 0.2697318196296692,
"learning_rate": 5.4043764052771e-06,
"loss": 1.3946113586425781,
"step": 228
},
{
"epoch": 1.0,
"grad_norm": 0.14553771913051605,
"learning_rate": 5.273573801905901e-06,
"loss": 1.330439805984497,
"step": 230
},
{
"epoch": 1.008695652173913,
"grad_norm": 0.16245104372501373,
"learning_rate": 5.152689034386213e-06,
"loss": 1.014844536781311,
"step": 232
},
{
"epoch": 1.017391304347826,
"grad_norm": 0.2378685176372528,
"learning_rate": 5.041822208981912e-06,
"loss": 0.32737967371940613,
"step": 234
},
{
"epoch": 1.0260869565217392,
"grad_norm": 0.20699624717235565,
"learning_rate": 4.941065135967408e-06,
"loss": 0.769230842590332,
"step": 236
},
{
"epoch": 1.0347826086956522,
"grad_norm": 0.12227105349302292,
"learning_rate": 4.850501253598363e-06,
"loss": 1.2364915609359741,
"step": 238
},
{
"epoch": 1.0434782608695652,
"grad_norm": 0.17606157064437866,
"learning_rate": 4.770205559015373e-06,
"loss": 1.2600164413452148,
"step": 240
},
{
"epoch": 1.0521739130434782,
"grad_norm": 0.24245694279670715,
"learning_rate": 4.700244546137851e-06,
"loss": 0.7779691815376282,
"step": 242
},
{
"epoch": 1.0608695652173914,
"grad_norm": 0.2091401368379593,
"learning_rate": 4.640676150599536e-06,
"loss": 1.0165550708770752,
"step": 244
},
{
"epoch": 1.0695652173913044,
"grad_norm": 0.12267211079597473,
"learning_rate": 4.5915497017712385e-06,
"loss": 1.0626590251922607,
"step": 246
},
{
"epoch": 1.0782608695652174,
"grad_norm": 0.1428578644990921,
"learning_rate": 4.552905881910531e-06,
"loss": 0.778892993927002,
"step": 248
},
{
"epoch": 1.0869565217391304,
"grad_norm": 0.12415514886379242,
"learning_rate": 4.524776692472253e-06,
"loss": 0.8480257391929626,
"step": 250
},
{
"epoch": 1.0956521739130434,
"grad_norm": 0.311179518699646,
"learning_rate": 4.507185427607667e-06,
"loss": 0.15872906148433685,
"step": 252
},
{
"epoch": 1.1043478260869566,
"grad_norm": 0.26307570934295654,
"learning_rate": 4.500146654874281e-06,
"loss": 1.3229949474334717,
"step": 254
},
{
"epoch": 1.1130434782608696,
"grad_norm": 0.12678803503513336,
"learning_rate": 4.50366620317225e-06,
"loss": 1.2751566171646118,
"step": 256
},
{
"epoch": 1.1217391304347826,
"grad_norm": 0.12393500655889511,
"learning_rate": 4.517741157917401e-06,
"loss": 0.7979919910430908,
"step": 258
},
{
"epoch": 1.1304347826086956,
"grad_norm": 0.11631522327661514,
"learning_rate": 4.542359863454825e-06,
"loss": 1.2487938404083252,
"step": 260
},
{
"epoch": 1.1391304347826088,
"grad_norm": 0.1201433464884758,
"learning_rate": 4.577501932711088e-06,
"loss": 1.503760814666748,
"step": 262
},
{
"epoch": 1.1478260869565218,
"grad_norm": 0.16790546476840973,
"learning_rate": 4.623138264077027e-06,
"loss": 1.349209189414978,
"step": 264
},
{
"epoch": 1.1565217391304348,
"grad_norm": 0.12929458916187286,
"learning_rate": 4.679231065507181e-06,
"loss": 1.3860739469528198,
"step": 266
},
{
"epoch": 1.1652173913043478,
"grad_norm": 1.2671191692352295,
"learning_rate": 4.745733885815875e-06,
"loss": 0.7751848101615906,
"step": 268
},
{
"epoch": 1.1739130434782608,
"grad_norm": 2.0448155403137207,
"learning_rate": 4.822591653144063e-06,
"loss": 0.8347904086112976,
"step": 270
},
{
"epoch": 1.182608695652174,
"grad_norm": 0.12114833295345306,
"learning_rate": 4.90974072056505e-06,
"loss": 1.281104564666748,
"step": 272
},
{
"epoch": 1.191304347826087,
"grad_norm": 0.1411694884300232,
"learning_rate": 5.0071089187913535e-06,
"loss": 1.2953712940216064,
"step": 274
},
{
"epoch": 1.2,
"grad_norm": 0.12283624708652496,
"learning_rate": 5.114615615939042e-06,
"loss": 1.0042020082473755,
"step": 276
},
{
"epoch": 1.208695652173913,
"grad_norm": 0.5199421048164368,
"learning_rate": 5.232171784300044e-06,
"loss": 0.8485981225967407,
"step": 278
},
{
"epoch": 1.2173913043478262,
"grad_norm": 0.25646281242370605,
"learning_rate": 5.359680074067178e-06,
"loss": 0.7609320282936096,
"step": 280
},
{
"epoch": 1.2260869565217392,
"grad_norm": 0.17641501128673553,
"learning_rate": 5.4970348939507965e-06,
"loss": 1.2442917823791504,
"step": 282
},
{
"epoch": 1.2347826086956522,
"grad_norm": 0.12078163027763367,
"learning_rate": 5.644122498620322e-06,
"loss": 0.7213406562805176,
"step": 284
},
{
"epoch": 1.2434782608695651,
"grad_norm": 0.2121657282114029,
"learning_rate": 5.800821082898257e-06,
"loss": 1.2524417638778687,
"step": 286
},
{
"epoch": 1.2521739130434781,
"grad_norm": 0.17032301425933838,
"learning_rate": 5.9670008826286325e-06,
"loss": 0.7535306215286255,
"step": 288
},
{
"epoch": 1.2608695652173914,
"grad_norm": 0.1459963470697403,
"learning_rate": 6.142524282136437e-06,
"loss": 1.0591073036193848,
"step": 290
},
{
"epoch": 1.2695652173913043,
"grad_norm": 0.10334640741348267,
"learning_rate": 6.327245928188936e-06,
"loss": 1.2697545289993286,
"step": 292
},
{
"epoch": 1.2782608695652173,
"grad_norm": 0.6387373805046082,
"learning_rate": 6.52101285036462e-06,
"loss": 0.8120895624160767,
"step": 294
},
{
"epoch": 1.2869565217391306,
"grad_norm": 0.09141097217798233,
"learning_rate": 6.7236645877299985e-06,
"loss": 0.16801220178604126,
"step": 296
},
{
"epoch": 1.2956521739130435,
"grad_norm": 0.14008603990077972,
"learning_rate": 6.935033321719419e-06,
"loss": 0.6829316020011902,
"step": 298
},
{
"epoch": 1.3043478260869565,
"grad_norm": 0.13336972892284393,
"learning_rate": 7.15494401510782e-06,
"loss": 0.24555981159210205,
"step": 300
},
{
"epoch": 1.3130434782608695,
"grad_norm": 0.15153667330741882,
"learning_rate": 7.383214556961368e-06,
"loss": 0.8401697874069214,
"step": 302
},
{
"epoch": 1.3217391304347825,
"grad_norm": 0.39388230443000793,
"learning_rate": 7.619655913445883e-06,
"loss": 1.6956262588500977,
"step": 304
},
{
"epoch": 1.3304347826086955,
"grad_norm": 0.0983978733420372,
"learning_rate": 7.864072284368284e-06,
"loss": 0.7533000111579895,
"step": 306
},
{
"epoch": 1.3391304347826087,
"grad_norm": 0.27145856618881226,
"learning_rate": 8.11626126532126e-06,
"loss": 1.0896899700164795,
"step": 308
},
{
"epoch": 1.3478260869565217,
"grad_norm": 0.16908416152000427,
"learning_rate": 8.376014015297057e-06,
"loss": 1.1746755838394165,
"step": 310
},
{
"epoch": 1.3565217391304347,
"grad_norm": 0.5621324777603149,
"learning_rate": 8.643115429631413e-06,
"loss": 0.28953975439071655,
"step": 312
},
{
"epoch": 1.365217391304348,
"grad_norm": 0.1343858242034912,
"learning_rate": 8.917344318134601e-06,
"loss": 1.2791374921798706,
"step": 314
},
{
"epoch": 1.373913043478261,
"grad_norm": 0.12179627269506454,
"learning_rate": 9.198473588261906e-06,
"loss": 1.5186599493026733,
"step": 316
},
{
"epoch": 1.382608695652174,
"grad_norm": 0.6460802555084229,
"learning_rate": 9.48627043317196e-06,
"loss": 0.7900709509849548,
"step": 318
},
{
"epoch": 1.391304347826087,
"grad_norm": 0.1162571832537651,
"learning_rate": 9.780496524517107e-06,
"loss": 1.5395351648330688,
"step": 320
},
{
"epoch": 1.4,
"grad_norm": 0.26111987233161926,
"learning_rate": 1.0080908209806292e-05,
"loss": 0.7073770761489868,
"step": 322
},
{
"epoch": 1.4086956521739131,
"grad_norm": 0.21174176037311554,
"learning_rate": 1.0387256714176855e-05,
"loss": 1.3696300983428955,
"step": 324
},
{
"epoch": 1.4173913043478261,
"grad_norm": 0.18966367840766907,
"learning_rate": 1.0699288346408284e-05,
"loss": 0.7893473505973816,
"step": 326
},
{
"epoch": 1.4260869565217391,
"grad_norm": 0.28794315457344055,
"learning_rate": 1.1016744709007203e-05,
"loss": 1.0100047588348389,
"step": 328
},
{
"epoch": 1.434782608695652,
"grad_norm": 0.26700359582901,
"learning_rate": 1.1339362912189803e-05,
"loss": 0.8350800275802612,
"step": 330
},
{
"epoch": 1.4434782608695653,
"grad_norm": 0.34453895688056946,
"learning_rate": 1.1666875791584241e-05,
"loss": 1.2977509498596191,
"step": 332
},
{
"epoch": 1.4521739130434783,
"grad_norm": 0.12350393086671829,
"learning_rate": 1.1999012129473044e-05,
"loss": 0.7434523105621338,
"step": 334
},
{
"epoch": 1.4608695652173913,
"grad_norm": 0.12331763654947281,
"learning_rate": 1.2335496879392006e-05,
"loss": 0.9966728091239929,
"step": 336
},
{
"epoch": 1.4695652173913043,
"grad_norm": 0.12660405039787292,
"learning_rate": 1.267605139389983e-05,
"loss": 1.2345221042633057,
"step": 338
},
{
"epoch": 1.4782608695652173,
"grad_norm": 0.5481416583061218,
"learning_rate": 1.3020393655329656e-05,
"loss": 0.4076097905635834,
"step": 340
},
{
"epoch": 1.4869565217391305,
"grad_norm": 0.1473967432975769,
"learning_rate": 1.3368238509331685e-05,
"loss": 1.0471733808517456,
"step": 342
},
{
"epoch": 1.4956521739130435,
"grad_norm": 0.14102469384670258,
"learning_rate": 1.3719297901013182e-05,
"loss": 1.2860357761383057,
"step": 344
},
{
"epoch": 1.5043478260869565,
"grad_norm": 0.29168060421943665,
"learning_rate": 1.4073281113480556e-05,
"loss": 1.0981141328811646,
"step": 346
},
{
"epoch": 1.5130434782608697,
"grad_norm": 0.12084399163722992,
"learning_rate": 1.4429895008585773e-05,
"loss": 1.5693000555038452,
"step": 348
},
{
"epoch": 1.5217391304347827,
"grad_norm": 0.3121376931667328,
"learning_rate": 1.4788844269678e-05,
"loss": 0.15966485440731049,
"step": 350
},
{
"epoch": 1.5304347826086957,
"grad_norm": 0.7553597092628479,
"learning_rate": 1.514983164615912e-05,
"loss": 0.7595433592796326,
"step": 352
},
{
"epoch": 1.5391304347826087,
"grad_norm": 0.4193861484527588,
"learning_rate": 1.5512558199640922e-05,
"loss": 0.8960363268852234,
"step": 354
},
{
"epoch": 1.5478260869565217,
"grad_norm": 2.7174839973449707,
"learning_rate": 1.587672355149983e-05,
"loss": 0.7889991998672485,
"step": 356
},
{
"epoch": 1.5565217391304347,
"grad_norm": 0.24500806629657745,
"learning_rate": 1.6242026131624477e-05,
"loss": 0.74623703956604,
"step": 358
},
{
"epoch": 1.5652173913043477,
"grad_norm": 0.12118079513311386,
"learning_rate": 1.6608163428149812e-05,
"loss": 1.2382371425628662,
"step": 360
},
{
"epoch": 1.5739130434782609,
"grad_norm": 0.21043698489665985,
"learning_rate": 1.697483223797118e-05,
"loss": 0.7253771424293518,
"step": 362
},
{
"epoch": 1.5826086956521739,
"grad_norm": 0.1428501456975937,
"learning_rate": 1.734172891783075e-05,
"loss": 0.6049482822418213,
"step": 364
},
{
"epoch": 1.591304347826087,
"grad_norm": 0.24034124612808228,
"learning_rate": 1.7708549635768553e-05,
"loss": 0.658636212348938,
"step": 366
},
{
"epoch": 1.6,
"grad_norm": 0.15072523057460785,
"learning_rate": 1.8074990622729676e-05,
"loss": 1.2467550039291382,
"step": 368
},
{
"epoch": 1.608695652173913,
"grad_norm": 0.29888659715652466,
"learning_rate": 1.8440748424119435e-05,
"loss": 0.332572340965271,
"step": 370
},
{
"epoch": 1.617391304347826,
"grad_norm": 0.4121609926223755,
"learning_rate": 1.8805520151098043e-05,
"loss": 0.7649952173233032,
"step": 372
},
{
"epoch": 1.626086956521739,
"grad_norm": 0.14216536283493042,
"learning_rate": 1.916900373140691e-05,
"loss": 0.8258912563323975,
"step": 374
},
{
"epoch": 1.634782608695652,
"grad_norm": 0.12106224149465561,
"learning_rate": 1.9530898159518646e-05,
"loss": 1.243992805480957,
"step": 376
},
{
"epoch": 1.643478260869565,
"grad_norm": 0.1799318641424179,
"learning_rate": 1.9890903745903662e-05,
"loss": 0.7369431853294373,
"step": 378
},
{
"epoch": 1.6521739130434783,
"grad_norm": 0.23679573833942413,
"learning_rate": 2.0248722365206985e-05,
"loss": 1.2425155639648438,
"step": 380
},
{
"epoch": 1.6608695652173913,
"grad_norm": 0.16253520548343658,
"learning_rate": 2.060405770312981e-05,
"loss": 0.8069741129875183,
"step": 382
},
{
"epoch": 1.6695652173913045,
"grad_norm": 0.12619397044181824,
"learning_rate": 2.0956615501811323e-05,
"loss": 0.7547456622123718,
"step": 384
},
{
"epoch": 1.6782608695652175,
"grad_norm": 0.2577739953994751,
"learning_rate": 2.1306103803507364e-05,
"loss": 1.0529491901397705,
"step": 386
},
{
"epoch": 1.6869565217391305,
"grad_norm": 0.32050907611846924,
"learning_rate": 2.1652233192364604e-05,
"loss": 1.4707834720611572,
"step": 388
},
{
"epoch": 1.6956521739130435,
"grad_norm": 0.1691839098930359,
"learning_rate": 2.1994717034089493e-05,
"loss": 0.7649831771850586,
"step": 390
},
{
"epoch": 1.7043478260869565,
"grad_norm": 0.11284446716308594,
"learning_rate": 2.2333271713314037e-05,
"loss": 1.237856388092041,
"step": 392
},
{
"epoch": 1.7130434782608694,
"grad_norm": 0.15681013464927673,
"learning_rate": 2.2667616868461238e-05,
"loss": 1.2670764923095703,
"step": 394
},
{
"epoch": 1.7217391304347827,
"grad_norm": 0.40354123711586,
"learning_rate": 2.299747562391633e-05,
"loss": 1.2364072799682617,
"step": 396
},
{
"epoch": 1.7304347826086957,
"grad_norm": 0.1340729296207428,
"learning_rate": 2.332257481931103e-05,
"loss": 0.7167130708694458,
"step": 398
},
{
"epoch": 1.7391304347826086,
"grad_norm": 0.11962085962295532,
"learning_rate": 2.3642645235731285e-05,
"loss": 0.7211905717849731,
"step": 400
},
{
"epoch": 1.7478260869565219,
"grad_norm": 0.1367277204990387,
"learning_rate": 2.3957421818660958e-05,
"loss": 0.7387893795967102,
"step": 402
},
{
"epoch": 1.7565217391304349,
"grad_norm": 0.12049642205238342,
"learning_rate": 2.4266643897476975e-05,
"loss": 1.2366782426834106,
"step": 404
},
{
"epoch": 1.7652173913043478,
"grad_norm": 0.19714173674583435,
"learning_rate": 2.457005540131405e-05,
"loss": 0.7809978127479553,
"step": 406
},
{
"epoch": 1.7739130434782608,
"grad_norm": 0.18713834881782532,
"learning_rate": 2.4867405071120378e-05,
"loss": 1.1861432790756226,
"step": 408
},
{
"epoch": 1.7826086956521738,
"grad_norm": 0.18612703680992126,
"learning_rate": 2.5158446667728563e-05,
"loss": 1.2835782766342163,
"step": 410
},
{
"epoch": 1.7913043478260868,
"grad_norm": 0.5197505354881287,
"learning_rate": 2.5442939175769457e-05,
"loss": 0.26566827297210693,
"step": 412
},
{
"epoch": 1.8,
"grad_norm": 0.16716937720775604,
"learning_rate": 2.5720647003260177e-05,
"loss": 1.253092646598816,
"step": 414
},
{
"epoch": 1.808695652173913,
"grad_norm": 0.16784952580928802,
"learning_rate": 2.5991340176700945e-05,
"loss": 1.298421859741211,
"step": 416
},
{
"epoch": 1.8173913043478263,
"grad_norm": 0.12985093891620636,
"learning_rate": 2.6254794531519122e-05,
"loss": 1.5727579593658447,
"step": 418
},
{
"epoch": 1.8260869565217392,
"grad_norm": 0.30004414916038513,
"learning_rate": 2.6510791897702884e-05,
"loss": 1.372650146484375,
"step": 420
},
{
"epoch": 1.8347826086956522,
"grad_norm": 0.16396737098693848,
"learning_rate": 2.675912028047062e-05,
"loss": 1.5493252277374268,
"step": 422
},
{
"epoch": 1.8434782608695652,
"grad_norm": 0.14453765749931335,
"learning_rate": 2.6999574035826618e-05,
"loss": 0.7752541899681091,
"step": 424
},
{
"epoch": 1.8521739130434782,
"grad_norm": 0.3888326585292816,
"learning_rate": 2.7231954040857604e-05,
"loss": 1.3314841985702515,
"step": 426
},
{
"epoch": 1.8608695652173912,
"grad_norm": 0.4329335391521454,
"learning_rate": 2.745606785862895e-05,
"loss": 0.36442431807518005,
"step": 428
},
{
"epoch": 1.8695652173913042,
"grad_norm": 0.19990986585617065,
"learning_rate": 2.767172989754432e-05,
"loss": 0.6660992503166199,
"step": 430
},
{
"epoch": 1.8782608695652174,
"grad_norm": 0.11110001057386398,
"learning_rate": 2.7878761565036436e-05,
"loss": 1.5394024848937988,
"step": 432
},
{
"epoch": 1.8869565217391304,
"grad_norm": 0.1121024340391159,
"learning_rate": 2.807699141546199e-05,
"loss": 0.7698279619216919,
"step": 434
},
{
"epoch": 1.8956521739130436,
"grad_norm": 0.15248660743236542,
"learning_rate": 2.826625529207797e-05,
"loss": 1.2329353094100952,
"step": 436
},
{
"epoch": 1.9043478260869566,
"grad_norm": 0.10831912606954575,
"learning_rate": 2.844639646298208e-05,
"loss": 1.2454890012741089,
"step": 438
},
{
"epoch": 1.9130434782608696,
"grad_norm": 0.1683008074760437,
"learning_rate": 2.8617265750904434e-05,
"loss": 1.2287697792053223,
"step": 440
},
{
"epoch": 1.9217391304347826,
"grad_norm": 0.08259912580251694,
"learning_rate": 2.8778721656743252e-05,
"loss": 0.7477931976318359,
"step": 442
},
{
"epoch": 1.9304347826086956,
"grad_norm": 0.11279872059822083,
"learning_rate": 2.893063047674209e-05,
"loss": 1.2804430723190308,
"step": 444
},
{
"epoch": 1.9391304347826086,
"grad_norm": 0.16371870040893555,
"learning_rate": 2.90728664132117e-05,
"loss": 0.7825669646263123,
"step": 446
},
{
"epoch": 1.9478260869565216,
"grad_norm": 0.08895144611597061,
"learning_rate": 2.9205311678704742e-05,
"loss": 0.6581708192825317,
"step": 448
},
{
"epoch": 1.9565217391304348,
"grad_norm": 0.3610806465148926,
"learning_rate": 2.932785659355714e-05,
"loss": 0.7231730222702026,
"step": 450
},
{
"epoch": 1.9652173913043478,
"grad_norm": 0.15365619957447052,
"learning_rate": 2.944039967671519e-05,
"loss": 0.7730036973953247,
"step": 452
},
{
"epoch": 1.973913043478261,
"grad_norm": 0.13084781169891357,
"learning_rate": 2.9542847729773423e-05,
"loss": 0.736638069152832,
"step": 454
},
{
"epoch": 1.982608695652174,
"grad_norm": 0.18188992142677307,
"learning_rate": 2.9635115914153388e-05,
"loss": 0.9782851934432983,
"step": 456
},
{
"epoch": 1.991304347826087,
"grad_norm": 0.18510757386684418,
"learning_rate": 2.971712782135961e-05,
"loss": 1.236344337463379,
"step": 458
},
{
"epoch": 2.0,
"grad_norm": 1.4175490140914917,
"learning_rate": 2.978881553625446e-05,
"loss": 1.1338064670562744,
"step": 460
},
{
"epoch": 2.008695652173913,
"grad_norm": 0.18492577970027924,
"learning_rate": 2.9850119693299577e-05,
"loss": 1.1988855600357056,
"step": 462
},
{
"epoch": 2.017391304347826,
"grad_norm": 0.24838964641094208,
"learning_rate": 2.9900989525717243e-05,
"loss": 0.1962333619594574,
"step": 464
},
{
"epoch": 2.026086956521739,
"grad_norm": 0.9311656951904297,
"learning_rate": 2.9941382907531037e-05,
"loss": 0.12543362379074097,
"step": 466
},
{
"epoch": 2.034782608695652,
"grad_norm": 0.22585399448871613,
"learning_rate": 2.9971266388450884e-05,
"loss": 0.7844517827033997,
"step": 468
},
{
"epoch": 2.0434782608695654,
"grad_norm": 0.29823383688926697,
"learning_rate": 2.9990615221573707e-05,
"loss": 0.7644298076629639,
"step": 470
},
{
"epoch": 2.0521739130434784,
"grad_norm": 0.1510343700647354,
"learning_rate": 2.999941338387663e-05,
"loss": 1.479148268699646,
"step": 472
},
{
"epoch": 2.0608695652173914,
"grad_norm": 0.7294149994850159,
"learning_rate": 2.999765358948589e-05,
"loss": 0.2585853636264801,
"step": 474
},
{
"epoch": 2.0695652173913044,
"grad_norm": 0.1448172628879547,
"learning_rate": 2.998533729571033e-05,
"loss": 1.2844960689544678,
"step": 476
},
{
"epoch": 2.0782608695652174,
"grad_norm": 0.25449809432029724,
"learning_rate": 2.9962474701834602e-05,
"loss": 0.7490485906600952,
"step": 478
},
{
"epoch": 2.0869565217391304,
"grad_norm": 0.1117057353258133,
"learning_rate": 2.9929084740673027e-05,
"loss": 1.1616085767745972,
"step": 480
},
{
"epoch": 2.0956521739130434,
"grad_norm": 0.5241631269454956,
"learning_rate": 2.9885195062891023e-05,
"loss": 1.7963777780532837,
"step": 482
},
{
"epoch": 2.1043478260869564,
"grad_norm": 0.16556978225708008,
"learning_rate": 2.9830842014107273e-05,
"loss": 1.2623823881149292,
"step": 484
},
{
"epoch": 2.1130434782608694,
"grad_norm": 0.5851399302482605,
"learning_rate": 2.9766070604795444e-05,
"loss": 0.681275486946106,
"step": 486
},
{
"epoch": 2.121739130434783,
"grad_norm": 0.4062120020389557,
"learning_rate": 2.969093447301039e-05,
"loss": 0.6947782039642334,
"step": 488
},
{
"epoch": 2.130434782608696,
"grad_norm": 0.14548048377037048,
"learning_rate": 2.9605495839969793e-05,
"loss": 0.6653367877006531,
"step": 490
},
{
"epoch": 2.139130434782609,
"grad_norm": 0.1378212869167328,
"learning_rate": 2.9509825458527955e-05,
"loss": 1.178207278251648,
"step": 492
},
{
"epoch": 2.1478260869565218,
"grad_norm": 0.5637624263763428,
"learning_rate": 2.940400255458444e-05,
"loss": 0.19959603250026703,
"step": 494
},
{
"epoch": 2.1565217391304348,
"grad_norm": 0.12706081569194794,
"learning_rate": 2.9288114761476036e-05,
"loss": 1.1562106609344482,
"step": 496
},
{
"epoch": 2.1652173913043478,
"grad_norm": 0.15716660022735596,
"learning_rate": 2.916225804740647e-05,
"loss": 1.1385471820831299,
"step": 498
},
{
"epoch": 2.1739130434782608,
"grad_norm": 0.11568231880664825,
"learning_rate": 2.9026536635973857e-05,
"loss": 0.49375131726264954,
"step": 500
},
{
"epoch": 2.1826086956521737,
"grad_norm": 0.16767719388008118,
"learning_rate": 2.888106291986178e-05,
"loss": 0.736272394657135,
"step": 502
},
{
"epoch": 2.1913043478260867,
"grad_norm": 0.1770874410867691,
"learning_rate": 2.872595736776543e-05,
"loss": 0.8101549744606018,
"step": 504
},
{
"epoch": 2.2,
"grad_norm": 0.8240585327148438,
"learning_rate": 2.8561348424629893e-05,
"loss": 0.2909260392189026,
"step": 506
},
{
"epoch": 2.208695652173913,
"grad_norm": 0.1195528656244278,
"learning_rate": 2.83873724052832e-05,
"loss": 0.7722839713096619,
"step": 508
},
{
"epoch": 2.217391304347826,
"grad_norm": 0.8490219712257385,
"learning_rate": 2.8204173381552252e-05,
"loss": 0.5638793110847473,
"step": 510
},
{
"epoch": 2.226086956521739,
"grad_norm": 0.3193603754043579,
"learning_rate": 2.8011903062955016e-05,
"loss": 0.31869935989379883,
"step": 512
},
{
"epoch": 2.234782608695652,
"grad_norm": 0.12175068259239197,
"learning_rate": 2.7810720671067893e-05,
"loss": 1.131230354309082,
"step": 514
},
{
"epoch": 2.243478260869565,
"grad_norm": 0.1269245743751526,
"learning_rate": 2.7600792807672248e-05,
"loss": 1.208681344985962,
"step": 516
},
{
"epoch": 2.252173913043478,
"grad_norm": 0.18002073466777802,
"learning_rate": 2.7382293316789273e-05,
"loss": 1.3903454542160034,
"step": 518
},
{
"epoch": 2.260869565217391,
"grad_norm": 0.6649693250656128,
"learning_rate": 2.7155403140717488e-05,
"loss": 0.7094644904136658,
"step": 520
},
{
"epoch": 2.269565217391304,
"grad_norm": 0.13027699291706085,
"learning_rate": 2.6920310170192024e-05,
"loss": 0.7558539509773254,
"step": 522
},
{
"epoch": 2.2782608695652176,
"grad_norm": 0.13726165890693665,
"learning_rate": 2.6677209088789823e-05,
"loss": 1.140393853187561,
"step": 524
},
{
"epoch": 2.2869565217391306,
"grad_norm": 0.21993081271648407,
"learning_rate": 2.642630121170968e-05,
"loss": 1.1117931604385376,
"step": 526
},
{
"epoch": 2.2956521739130435,
"grad_norm": 0.1521347314119339,
"learning_rate": 2.616779431906042e-05,
"loss": 0.6197381019592285,
"step": 528
},
{
"epoch": 2.3043478260869565,
"grad_norm": 0.17388762533664703,
"learning_rate": 2.5901902483795464e-05,
"loss": 1.1404460668563843,
"step": 530
},
{
"epoch": 2.3130434782608695,
"grad_norm": 0.19341905415058136,
"learning_rate": 2.5628845894436232e-05,
"loss": 1.110768437385559,
"step": 532
},
{
"epoch": 2.3217391304347825,
"grad_norm": 0.11132687330245972,
"learning_rate": 2.5348850672731114e-05,
"loss": 0.18440936505794525,
"step": 534
},
{
"epoch": 2.3304347826086955,
"grad_norm": 0.12755344808101654,
"learning_rate": 2.506214868640111e-05,
"loss": 0.6152647137641907,
"step": 536
},
{
"epoch": 2.3391304347826085,
"grad_norm": 0.07117603719234467,
"learning_rate": 2.4768977357127068e-05,
"loss": 0.6943525075912476,
"step": 538
},
{
"epoch": 2.3478260869565215,
"grad_norm": 0.22992900013923645,
"learning_rate": 2.4469579463937736e-05,
"loss": 1.155451774597168,
"step": 540
},
{
"epoch": 2.356521739130435,
"grad_norm": 0.13336700201034546,
"learning_rate": 2.4164202942161202e-05,
"loss": 0.6562126874923706,
"step": 542
},
{
"epoch": 2.365217391304348,
"grad_norm": 0.17184697091579437,
"learning_rate": 2.385310067810646e-05,
"loss": 1.1728577613830566,
"step": 544
},
{
"epoch": 2.373913043478261,
"grad_norm": 0.24630305171012878,
"learning_rate": 2.353653029964486e-05,
"loss": 0.7028173208236694,
"step": 546
},
{
"epoch": 2.382608695652174,
"grad_norm": 0.4979984164237976,
"learning_rate": 2.3214753962865128e-05,
"loss": 1.0240296125411987,
"step": 548
},
{
"epoch": 2.391304347826087,
"grad_norm": 0.786608099937439,
"learning_rate": 2.2888038134978446e-05,
"loss": 0.6851329207420349,
"step": 550
},
{
"epoch": 2.4,
"grad_norm": 0.21292582154273987,
"learning_rate": 2.2556653373653506e-05,
"loss": 1.1324046850204468,
"step": 552
},
{
"epoch": 2.408695652173913,
"grad_norm": 0.07219722867012024,
"learning_rate": 2.2220874102964066e-05,
"loss": 0.5204906463623047,
"step": 554
},
{
"epoch": 2.417391304347826,
"grad_norm": 0.37272965908050537,
"learning_rate": 2.1880978386134822e-05,
"loss": 1.3587431907653809,
"step": 556
},
{
"epoch": 2.426086956521739,
"grad_norm": 0.1551341414451599,
"learning_rate": 2.1537247695273588e-05,
"loss": 0.15599098801612854,
"step": 558
},
{
"epoch": 2.4347826086956523,
"grad_norm": 0.10612637549638748,
"learning_rate": 2.1189966678280585e-05,
"loss": 1.1518821716308594,
"step": 560
},
{
"epoch": 2.4434782608695653,
"grad_norm": 0.2193143665790558,
"learning_rate": 2.0839422923127686e-05,
"loss": 0.5007261037826538,
"step": 562
},
{
"epoch": 2.4521739130434783,
"grad_norm": 0.45728591084480286,
"learning_rate": 2.0485906719703126e-05,
"loss": 1.1414357423782349,
"step": 564
},
{
"epoch": 2.4608695652173913,
"grad_norm": 0.19980177283287048,
"learning_rate": 2.0129710819418574e-05,
"loss": 0.6164332032203674,
"step": 566
},
{
"epoch": 2.4695652173913043,
"grad_norm": 0.12282229214906693,
"learning_rate": 1.9771130192777892e-05,
"loss": 1.0142958164215088,
"step": 568
},
{
"epoch": 2.4782608695652173,
"grad_norm": 0.4857296347618103,
"learning_rate": 1.9410461785108178e-05,
"loss": 0.7631978392601013,
"step": 570
},
{
"epoch": 2.4869565217391303,
"grad_norm": 0.2022370547056198,
"learning_rate": 1.9048004270655354e-05,
"loss": 1.2384475469589233,
"step": 572
},
{
"epoch": 2.4956521739130437,
"grad_norm": 1.394151210784912,
"learning_rate": 1.868405780524824e-05,
"loss": 0.38715630769729614,
"step": 574
},
{
"epoch": 2.5043478260869563,
"grad_norm": 0.1383918970823288,
"learning_rate": 1.831892377773547e-05,
"loss": 1.1841545104980469,
"step": 576
},
{
"epoch": 2.5130434782608697,
"grad_norm": 0.19289281964302063,
"learning_rate": 1.795290456040148e-05,
"loss": 1.3352640867233276,
"step": 578
},
{
"epoch": 2.5217391304347827,
"grad_norm": 0.26483142375946045,
"learning_rate": 1.7586303258567964e-05,
"loss": 0.6390936374664307,
"step": 580
},
{
"epoch": 2.5304347826086957,
"grad_norm": 0.4453805088996887,
"learning_rate": 1.7219423459588513e-05,
"loss": 0.9683659672737122,
"step": 582
},
{
"epoch": 2.5391304347826087,
"grad_norm": 0.12072495371103287,
"learning_rate": 1.6852568981443806e-05,
"loss": 1.1052292585372925,
"step": 584
},
{
"epoch": 2.5478260869565217,
"grad_norm": 0.13637158274650574,
"learning_rate": 1.648604362114606e-05,
"loss": 0.708678126335144,
"step": 586
},
{
"epoch": 2.5565217391304347,
"grad_norm": 0.10113801807165146,
"learning_rate": 1.6120150903160605e-05,
"loss": 0.06356725841760635,
"step": 588
},
{
"epoch": 2.5652173913043477,
"grad_norm": 0.1465790867805481,
"learning_rate": 1.5755193828053435e-05,
"loss": 0.694472074508667,
"step": 590
},
{
"epoch": 2.573913043478261,
"grad_norm": 0.4859929382801056,
"learning_rate": 1.539147462157235e-05,
"loss": 0.09391395002603531,
"step": 592
},
{
"epoch": 2.5826086956521737,
"grad_norm": 0.4653010368347168,
"learning_rate": 1.5029294484369874e-05,
"loss": 1.3817299604415894,
"step": 594
},
{
"epoch": 2.591304347826087,
"grad_norm": 0.23374710977077484,
"learning_rate": 1.4668953342575074e-05,
"loss": 0.28024399280548096,
"step": 596
},
{
"epoch": 2.6,
"grad_norm": 0.19562578201293945,
"learning_rate": 1.4310749599420711e-05,
"loss": 0.4380612373352051,
"step": 598
},
{
"epoch": 2.608695652173913,
"grad_norm": 0.11721507459878922,
"learning_rate": 1.3954979888131745e-05,
"loss": 1.1383816003799438,
"step": 600
},
{
"epoch": 2.617391304347826,
"grad_norm": 0.12313393503427505,
"learning_rate": 1.3601938826279298e-05,
"loss": 0.6719468235969543,
"step": 602
},
{
"epoch": 2.626086956521739,
"grad_norm": 0.2674733102321625,
"learning_rate": 1.325191877180408e-05,
"loss": 0.8287115097045898,
"step": 604
},
{
"epoch": 2.634782608695652,
"grad_norm": 0.1634570211172104,
"learning_rate": 1.2905209580910897e-05,
"loss": 0.6804332137107849,
"step": 606
},
{
"epoch": 2.643478260869565,
"grad_norm": 0.2214396893978119,
"learning_rate": 1.2562098368034954e-05,
"loss": 1.1206378936767578,
"step": 608
},
{
"epoch": 2.6521739130434785,
"grad_norm": 0.17025478184223175,
"learning_rate": 1.2222869268078575e-05,
"loss": 0.7674025297164917,
"step": 610
},
{
"epoch": 2.660869565217391,
"grad_norm": 0.17491810023784637,
"learning_rate": 1.1887803201115396e-05,
"loss": 1.1401374340057373,
"step": 612
},
{
"epoch": 2.6695652173913045,
"grad_norm": 0.12606146931648254,
"learning_rate": 1.1557177639756725e-05,
"loss": 1.3836885690689087,
"step": 614
},
{
"epoch": 2.6782608695652175,
"grad_norm": 0.1093195304274559,
"learning_rate": 1.1231266379372912e-05,
"loss": 0.737335205078125,
"step": 616
},
{
"epoch": 2.6869565217391305,
"grad_norm": 0.14942651987075806,
"learning_rate": 1.0910339311359677e-05,
"loss": 0.6436473727226257,
"step": 618
},
{
"epoch": 2.6956521739130435,
"grad_norm": 0.6913372874259949,
"learning_rate": 1.0594662199637597e-05,
"loss": 0.7471338510513306,
"step": 620
},
{
"epoch": 2.7043478260869565,
"grad_norm": 0.14125873148441315,
"learning_rate": 1.028449646056948e-05,
"loss": 1.1643157005310059,
"step": 622
},
{
"epoch": 2.7130434782608694,
"grad_norm": 0.13380275666713715,
"learning_rate": 9.980098946478024e-06,
"loss": 1.1436272859573364,
"step": 624
},
{
"epoch": 2.7217391304347824,
"grad_norm": 0.1280951350927353,
"learning_rate": 9.68172173294313e-06,
"loss": 1.134834885597229,
"step": 626
},
{
"epoch": 2.730434782608696,
"grad_norm": 0.15834616124629974,
"learning_rate": 9.389611910054602e-06,
"loss": 1.237627625465393,
"step": 628
},
{
"epoch": 2.7391304347826084,
"grad_norm": 0.20524626970291138,
"learning_rate": 9.10401137779385e-06,
"loss": 1.1352347135543823,
"step": 630
},
{
"epoch": 2.747826086956522,
"grad_norm": 0.6567396521568298,
"learning_rate": 8.825156645713136e-06,
"loss": 0.46068328619003296,
"step": 632
},
{
"epoch": 2.756521739130435,
"grad_norm": 0.11645974218845367,
"learning_rate": 8.553278637079007e-06,
"loss": 0.6736262440681458,
"step": 634
},
{
"epoch": 2.765217391304348,
"grad_norm": 0.14792583882808685,
"learning_rate": 8.28860249764168e-06,
"loss": 0.6648976802825928,
"step": 636
},
{
"epoch": 2.773913043478261,
"grad_norm": 0.17419379949569702,
"learning_rate": 8.031347409188885e-06,
"loss": 1.132703423500061,
"step": 638
},
{
"epoch": 2.782608695652174,
"grad_norm": 0.323537677526474,
"learning_rate": 7.78172640803854e-06,
"loss": 1.2038984298706055,
"step": 640
},
{
"epoch": 2.791304347826087,
"grad_norm": 0.2815156877040863,
"learning_rate": 7.539946208620638e-06,
"loss": 1.206749677658081,
"step": 642
},
{
"epoch": 2.8,
"grad_norm": 0.5039319396018982,
"learning_rate": 7.306207032294216e-06,
"loss": 0.7191698551177979,
"step": 644
},
{
"epoch": 2.8086956521739133,
"grad_norm": 0.18908892571926117,
"learning_rate": 7.08070244154148e-06,
"loss": 0.7605520486831665,
"step": 646
},
{
"epoch": 2.8173913043478263,
"grad_norm": 0.2733303904533386,
"learning_rate": 6.8636191796761585e-06,
"loss": 0.6928582787513733,
"step": 648
},
{
"epoch": 2.8260869565217392,
"grad_norm": 0.30275958776474,
"learning_rate": 6.655137016198907e-06,
"loss": 1.3508784770965576,
"step": 650
},
{
"epoch": 2.8347826086956522,
"grad_norm": 0.1958470344543457,
"learning_rate": 6.455428597927829e-06,
"loss": 0.9979486465454102,
"step": 652
},
{
"epoch": 2.8434782608695652,
"grad_norm": 0.14218895137310028,
"learning_rate": 6.264659306027376e-06,
"loss": 1.139778971672058,
"step": 654
},
{
"epoch": 2.8521739130434782,
"grad_norm": 0.17221413552761078,
"learning_rate": 6.082987119054014e-06,
"loss": 0.8573166728019714,
"step": 656
},
{
"epoch": 2.860869565217391,
"grad_norm": 0.13826826214790344,
"learning_rate": 5.910562482132162e-06,
"loss": 1.1786668300628662,
"step": 658
},
{
"epoch": 2.869565217391304,
"grad_norm": 0.3460427522659302,
"learning_rate": 5.747528182368537e-06,
"loss": 1.1484174728393555,
"step": 660
},
{
"epoch": 2.878260869565217,
"grad_norm": 0.21556347608566284,
"learning_rate": 5.594019230608316e-06,
"loss": 1.2150859832763672,
"step": 662
},
{
"epoch": 2.8869565217391306,
"grad_norm": 0.17000877857208252,
"learning_rate": 5.4501627496308754e-06,
"loss": 1.1722112894058228,
"step": 664
},
{
"epoch": 2.8956521739130436,
"grad_norm": 0.1485866904258728,
"learning_rate": 5.316077868877737e-06,
"loss": 1.1153019666671753,
"step": 666
},
{
"epoch": 2.9043478260869566,
"grad_norm": 0.12437450885772705,
"learning_rate": 5.191875625799956e-06,
"loss": 1.164955496788025,
"step": 668
},
{
"epoch": 2.9130434782608696,
"grad_norm": 0.14834971725940704,
"learning_rate": 5.077658873906456e-06,
"loss": 1.1394230127334595,
"step": 670
},
{
"epoch": 2.9217391304347826,
"grad_norm": 1.5700373649597168,
"learning_rate": 4.973522197589804e-06,
"loss": 0.6004053354263306,
"step": 672
},
{
"epoch": 2.9304347826086956,
"grad_norm": 0.3607974648475647,
"learning_rate": 4.87955183379954e-06,
"loss": 0.8902421593666077,
"step": 674
},
{
"epoch": 2.9391304347826086,
"grad_norm": 0.12654371559619904,
"learning_rate": 4.795825600628273e-06,
"loss": 1.1788089275360107,
"step": 676
},
{
"epoch": 2.9478260869565216,
"grad_norm": 0.18413525819778442,
"learning_rate": 4.722412832869478e-06,
"loss": 1.0090527534484863,
"step": 678
},
{
"epoch": 2.9565217391304346,
"grad_norm": 0.13298968970775604,
"learning_rate": 4.659374324600457e-06,
"loss": 1.156622290611267,
"step": 680
},
{
"epoch": 2.965217391304348,
"grad_norm": 0.27358752489089966,
"learning_rate": 4.606762278837947e-06,
"loss": 0.44529619812965393,
"step": 682
},
{
"epoch": 2.973913043478261,
"grad_norm": 0.2820883095264435,
"learning_rate": 4.5646202643081405e-06,
"loss": 0.7129748463630676,
"step": 684
},
{
"epoch": 2.982608695652174,
"grad_norm": 0.09756142646074295,
"learning_rate": 4.532983179366819e-06,
"loss": 0.8244656920433044,
"step": 686
},
{
"epoch": 2.991304347826087,
"grad_norm": 0.19994956254959106,
"learning_rate": 4.511877223099601e-06,
"loss": 1.285622477531433,
"step": 688
},
{
"epoch": 3.0,
"grad_norm": 0.058878399431705475,
"learning_rate": 4.501319873626102e-06,
"loss": 0.6546279191970825,
"step": 690
},
{
"epoch": 3.0,
"step": 690,
"total_flos": 2.5683697677065257e+18,
"train_loss": 1.0660493817018426,
"train_runtime": 18433.9416,
"train_samples_per_second": 2.396,
"train_steps_per_second": 0.037
}
],
"logging_steps": 2,
"max_steps": 690,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 99999,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.5683697677065257e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}