imdatta0's picture
End of training
8b1b653 verified
raw
history blame
27.2 kB
{
"best_metric": 2.017824411392212,
"best_model_checkpoint": "/home/datta0/models/lora_final/Mistral-7B-v0.3_pct_default_r16/checkpoint-376",
"epoch": 0.969931479242241,
"eval_steps": 8,
"global_step": 376,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0025796049979846837,
"grad_norm": 16.779726028442383,
"learning_rate": 1.25e-05,
"loss": 2.1524,
"step": 1
},
{
"epoch": 0.010318419991938735,
"grad_norm": 10.524264335632324,
"learning_rate": 5e-05,
"loss": 2.0461,
"step": 4
},
{
"epoch": 0.02063683998387747,
"grad_norm": 10.28955078125,
"learning_rate": 0.0001,
"loss": 1.9646,
"step": 8
},
{
"epoch": 0.02063683998387747,
"eval_loss": 2.0348494052886963,
"eval_runtime": 136.7937,
"eval_samples_per_second": 1.791,
"eval_steps_per_second": 0.899,
"step": 8
},
{
"epoch": 0.030955259975816204,
"grad_norm": 19.959510803222656,
"learning_rate": 9.997251843068762e-05,
"loss": 2.012,
"step": 12
},
{
"epoch": 0.04127367996775494,
"grad_norm": 14.419052124023438,
"learning_rate": 9.989010393221656e-05,
"loss": 2.0531,
"step": 16
},
{
"epoch": 0.04127367996775494,
"eval_loss": 2.034454107284546,
"eval_runtime": 132.3292,
"eval_samples_per_second": 1.851,
"eval_steps_per_second": 0.93,
"step": 16
},
{
"epoch": 0.051592099959693674,
"grad_norm": 17.94999122619629,
"learning_rate": 9.97528470997769e-05,
"loss": 2.1603,
"step": 20
},
{
"epoch": 0.06191051995163241,
"grad_norm": 14.041346549987793,
"learning_rate": 9.956089881469482e-05,
"loss": 2.1168,
"step": 24
},
{
"epoch": 0.06191051995163241,
"eval_loss": 2.046419382095337,
"eval_runtime": 124.8023,
"eval_samples_per_second": 1.963,
"eval_steps_per_second": 0.986,
"step": 24
},
{
"epoch": 0.07222893994357114,
"grad_norm": 16.775529861450195,
"learning_rate": 9.931447007857432e-05,
"loss": 2.075,
"step": 28
},
{
"epoch": 0.08254735993550988,
"grad_norm": 16.405385971069336,
"learning_rate": 9.901383178135113e-05,
"loss": 2.0712,
"step": 32
},
{
"epoch": 0.08254735993550988,
"eval_loss": 2.046201705932617,
"eval_runtime": 241.0686,
"eval_samples_per_second": 1.016,
"eval_steps_per_second": 0.51,
"step": 32
},
{
"epoch": 0.09286577992744861,
"grad_norm": 27.64794158935547,
"learning_rate": 9.865931440351337e-05,
"loss": 2.0807,
"step": 36
},
{
"epoch": 0.10318419991938735,
"grad_norm": 20.194869995117188,
"learning_rate": 9.825130765281668e-05,
"loss": 2.0779,
"step": 40
},
{
"epoch": 0.10318419991938735,
"eval_loss": 2.0509941577911377,
"eval_runtime": 141.7947,
"eval_samples_per_second": 1.728,
"eval_steps_per_second": 0.867,
"step": 40
},
{
"epoch": 0.11350261991132608,
"grad_norm": 18.115293502807617,
"learning_rate": 9.779026003589304e-05,
"loss": 2.065,
"step": 44
},
{
"epoch": 0.12382103990326482,
"grad_norm": 14.515559196472168,
"learning_rate": 9.727667836522407e-05,
"loss": 2.0905,
"step": 48
},
{
"epoch": 0.12382103990326482,
"eval_loss": 2.0475986003875732,
"eval_runtime": 128.356,
"eval_samples_per_second": 1.909,
"eval_steps_per_second": 0.958,
"step": 48
},
{
"epoch": 0.13413945989520354,
"grad_norm": 15.200082778930664,
"learning_rate": 9.6711127202021e-05,
"loss": 2.049,
"step": 52
},
{
"epoch": 0.1444578798871423,
"grad_norm": 14.234698295593262,
"learning_rate": 9.609422823562345e-05,
"loss": 2.0624,
"step": 56
},
{
"epoch": 0.1444578798871423,
"eval_loss": 2.0475313663482666,
"eval_runtime": 123.5017,
"eval_samples_per_second": 1.984,
"eval_steps_per_second": 0.996,
"step": 56
},
{
"epoch": 0.154776299879081,
"grad_norm": 34.070438385009766,
"learning_rate": 9.542665960009959e-05,
"loss": 2.0794,
"step": 60
},
{
"epoch": 0.16509471987101976,
"grad_norm": 14.070226669311523,
"learning_rate": 9.470915512879852e-05,
"loss": 2.0793,
"step": 64
},
{
"epoch": 0.16509471987101976,
"eval_loss": 2.042788505554199,
"eval_runtime": 124.8794,
"eval_samples_per_second": 1.962,
"eval_steps_per_second": 0.985,
"step": 64
},
{
"epoch": 0.17541313986295848,
"grad_norm": 13.649319648742676,
"learning_rate": 9.394250354767467e-05,
"loss": 2.0942,
"step": 68
},
{
"epoch": 0.18573155985489723,
"grad_norm": 14.56338882446289,
"learning_rate": 9.312754760827061e-05,
"loss": 2.0559,
"step": 72
},
{
"epoch": 0.18573155985489723,
"eval_loss": 2.052072763442993,
"eval_runtime": 252.4609,
"eval_samples_per_second": 0.97,
"eval_steps_per_second": 0.487,
"step": 72
},
{
"epoch": 0.19604997984683595,
"grad_norm": 16.887882232666016,
"learning_rate": 9.226518316131176e-05,
"loss": 2.0926,
"step": 76
},
{
"epoch": 0.2063683998387747,
"grad_norm": 16.470928192138672,
"learning_rate": 9.1356358171931e-05,
"loss": 2.0597,
"step": 80
},
{
"epoch": 0.2063683998387747,
"eval_loss": 2.068044424057007,
"eval_runtime": 120.7858,
"eval_samples_per_second": 2.028,
"eval_steps_per_second": 1.018,
"step": 80
},
{
"epoch": 0.21668681983071342,
"grad_norm": 16.528242111206055,
"learning_rate": 9.040207167760586e-05,
"loss": 2.0819,
"step": 84
},
{
"epoch": 0.22700523982265217,
"grad_norm": 18.425939559936523,
"learning_rate": 8.940337268995385e-05,
"loss": 2.1235,
"step": 88
},
{
"epoch": 0.22700523982265217,
"eval_loss": 2.0848894119262695,
"eval_runtime": 117.9661,
"eval_samples_per_second": 2.077,
"eval_steps_per_second": 1.043,
"step": 88
},
{
"epoch": 0.2373236598145909,
"grad_norm": 11.150261878967285,
"learning_rate": 8.836135904159302e-05,
"loss": 2.1182,
"step": 92
},
{
"epoch": 0.24764207980652964,
"grad_norm": 13.942819595336914,
"learning_rate": 8.727717617933544e-05,
"loss": 2.14,
"step": 96
},
{
"epoch": 0.24764207980652964,
"eval_loss": 2.0771827697753906,
"eval_runtime": 147.6689,
"eval_samples_per_second": 1.659,
"eval_steps_per_second": 0.833,
"step": 96
},
{
"epoch": 0.25796049979846836,
"grad_norm": 21.486312866210938,
"learning_rate": 8.615201590504017e-05,
"loss": 2.1272,
"step": 100
},
{
"epoch": 0.2682789197904071,
"grad_norm": 13.030159950256348,
"learning_rate": 8.498711506550983e-05,
"loss": 2.1586,
"step": 104
},
{
"epoch": 0.2682789197904071,
"eval_loss": 2.0880281925201416,
"eval_runtime": 155.8186,
"eval_samples_per_second": 1.572,
"eval_steps_per_second": 0.789,
"step": 104
},
{
"epoch": 0.27859733978234585,
"grad_norm": 18.889177322387695,
"learning_rate": 8.378375419287099e-05,
"loss": 2.0883,
"step": 108
},
{
"epoch": 0.2889157597742846,
"grad_norm": 16.2430419921875,
"learning_rate": 8.25432560969328e-05,
"loss": 2.0974,
"step": 112
},
{
"epoch": 0.2889157597742846,
"eval_loss": 2.0836544036865234,
"eval_runtime": 249.1613,
"eval_samples_per_second": 0.983,
"eval_steps_per_second": 0.494,
"step": 112
},
{
"epoch": 0.2992341797662233,
"grad_norm": 16.362117767333984,
"learning_rate": 8.126698441107146e-05,
"loss": 2.1367,
"step": 116
},
{
"epoch": 0.309552599758162,
"grad_norm": 16.559032440185547,
"learning_rate": 7.995634209323886e-05,
"loss": 2.1577,
"step": 120
},
{
"epoch": 0.309552599758162,
"eval_loss": 2.0837860107421875,
"eval_runtime": 142.7033,
"eval_samples_per_second": 1.717,
"eval_steps_per_second": 0.862,
"step": 120
},
{
"epoch": 0.3198710197501008,
"grad_norm": 16.882129669189453,
"learning_rate": 7.861276988374302e-05,
"loss": 2.1391,
"step": 124
},
{
"epoch": 0.3301894397420395,
"grad_norm": 16.510204315185547,
"learning_rate": 7.723774472149601e-05,
"loss": 2.0998,
"step": 128
},
{
"epoch": 0.3301894397420395,
"eval_loss": 2.089890718460083,
"eval_runtime": 134.7898,
"eval_samples_per_second": 1.818,
"eval_steps_per_second": 0.913,
"step": 128
},
{
"epoch": 0.34050785973397824,
"grad_norm": 15.429149627685547,
"learning_rate": 7.583277812046993e-05,
"loss": 2.0801,
"step": 132
},
{
"epoch": 0.35082627972591696,
"grad_norm": 16.833860397338867,
"learning_rate": 7.439941450814591e-05,
"loss": 2.1069,
"step": 136
},
{
"epoch": 0.35082627972591696,
"eval_loss": 2.088224172592163,
"eval_runtime": 119.5762,
"eval_samples_per_second": 2.049,
"eval_steps_per_second": 1.029,
"step": 136
},
{
"epoch": 0.3611446997178557,
"grad_norm": 15.170039176940918,
"learning_rate": 7.293922952778239e-05,
"loss": 2.1277,
"step": 140
},
{
"epoch": 0.37146311970979445,
"grad_norm": 16.08381462097168,
"learning_rate": 7.145382830636924e-05,
"loss": 2.1621,
"step": 144
},
{
"epoch": 0.37146311970979445,
"eval_loss": 2.084606647491455,
"eval_runtime": 51.8689,
"eval_samples_per_second": 4.723,
"eval_steps_per_second": 2.371,
"step": 144
},
{
"epoch": 0.3817815397017332,
"grad_norm": 13.434906005859375,
"learning_rate": 6.994484369017143e-05,
"loss": 2.124,
"step": 148
},
{
"epoch": 0.3920999596936719,
"grad_norm": 16.64518165588379,
"learning_rate": 6.841393444980177e-05,
"loss": 2.1441,
"step": 152
},
{
"epoch": 0.3920999596936719,
"eval_loss": 2.0949418544769287,
"eval_runtime": 54.9018,
"eval_samples_per_second": 4.463,
"eval_steps_per_second": 2.24,
"step": 152
},
{
"epoch": 0.4024183796856106,
"grad_norm": 14.433173179626465,
"learning_rate": 6.686278345679625e-05,
"loss": 2.1492,
"step": 156
},
{
"epoch": 0.4127367996775494,
"grad_norm": 21.335298538208008,
"learning_rate": 6.529309583369605e-05,
"loss": 2.1355,
"step": 160
},
{
"epoch": 0.4127367996775494,
"eval_loss": 2.085907220840454,
"eval_runtime": 31.9309,
"eval_samples_per_second": 7.673,
"eval_steps_per_second": 3.852,
"step": 160
},
{
"epoch": 0.4230552196694881,
"grad_norm": 12.250011444091797,
"learning_rate": 6.370659707966967e-05,
"loss": 2.1237,
"step": 164
},
{
"epoch": 0.43337363966142683,
"grad_norm": 16.425058364868164,
"learning_rate": 6.2105031173736e-05,
"loss": 2.084,
"step": 168
},
{
"epoch": 0.43337363966142683,
"eval_loss": 2.087137460708618,
"eval_runtime": 33.4112,
"eval_samples_per_second": 7.333,
"eval_steps_per_second": 3.681,
"step": 168
},
{
"epoch": 0.44369205965336556,
"grad_norm": 13.831182479858398,
"learning_rate": 6.049015865767318e-05,
"loss": 2.138,
"step": 172
},
{
"epoch": 0.45401047964530433,
"grad_norm": 17.53313446044922,
"learning_rate": 5.88637547007204e-05,
"loss": 2.1649,
"step": 176
},
{
"epoch": 0.45401047964530433,
"eval_loss": 2.084533214569092,
"eval_runtime": 32.4375,
"eval_samples_per_second": 7.553,
"eval_steps_per_second": 3.792,
"step": 176
},
{
"epoch": 0.46432889963724305,
"grad_norm": 14.465840339660645,
"learning_rate": 5.722760714820057e-05,
"loss": 2.1365,
"step": 180
},
{
"epoch": 0.4746473196291818,
"grad_norm": 14.579000473022461,
"learning_rate": 5.5583514556208514e-05,
"loss": 2.0651,
"step": 184
},
{
"epoch": 0.4746473196291818,
"eval_loss": 2.071864604949951,
"eval_runtime": 51.0877,
"eval_samples_per_second": 4.796,
"eval_steps_per_second": 2.408,
"step": 184
},
{
"epoch": 0.4849657396211205,
"grad_norm": 15.363332748413086,
"learning_rate": 5.393328421452514e-05,
"loss": 2.1009,
"step": 188
},
{
"epoch": 0.49528415961305927,
"grad_norm": 13.905656814575195,
"learning_rate": 5.2278730159931076e-05,
"loss": 2.1708,
"step": 192
},
{
"epoch": 0.49528415961305927,
"eval_loss": 2.072157144546509,
"eval_runtime": 32.8055,
"eval_samples_per_second": 7.468,
"eval_steps_per_second": 3.749,
"step": 192
},
{
"epoch": 0.505602579604998,
"grad_norm": 13.91466236114502,
"learning_rate": 5.062167118210367e-05,
"loss": 2.1383,
"step": 196
},
{
"epoch": 0.5159209995969367,
"grad_norm": 15.01607608795166,
"learning_rate": 4.896392882428901e-05,
"loss": 2.1311,
"step": 200
},
{
"epoch": 0.5159209995969367,
"eval_loss": 2.0677082538604736,
"eval_runtime": 33.5658,
"eval_samples_per_second": 7.299,
"eval_steps_per_second": 3.664,
"step": 200
},
{
"epoch": 0.5262394195888754,
"grad_norm": 10.953645706176758,
"learning_rate": 4.730732538094749e-05,
"loss": 2.1166,
"step": 204
},
{
"epoch": 0.5365578395808142,
"grad_norm": 15.371764183044434,
"learning_rate": 4.565368189457313e-05,
"loss": 2.1038,
"step": 208
},
{
"epoch": 0.5365578395808142,
"eval_loss": 2.0626721382141113,
"eval_runtime": 32.7774,
"eval_samples_per_second": 7.475,
"eval_steps_per_second": 3.753,
"step": 208
},
{
"epoch": 0.5468762595727529,
"grad_norm": 14.383162498474121,
"learning_rate": 4.400481615388948e-05,
"loss": 2.1023,
"step": 212
},
{
"epoch": 0.5571946795646917,
"grad_norm": 15.261186599731445,
"learning_rate": 4.236254069562213e-05,
"loss": 2.0804,
"step": 216
},
{
"epoch": 0.5571946795646917,
"eval_loss": 2.075681686401367,
"eval_runtime": 32.6725,
"eval_samples_per_second": 7.499,
"eval_steps_per_second": 3.765,
"step": 216
},
{
"epoch": 0.5675130995566304,
"grad_norm": 10.065788269042969,
"learning_rate": 4.0728660812044536e-05,
"loss": 2.0796,
"step": 220
},
{
"epoch": 0.5778315195485692,
"grad_norm": 13.413424491882324,
"learning_rate": 3.910497256648742e-05,
"loss": 2.0695,
"step": 224
},
{
"epoch": 0.5778315195485692,
"eval_loss": 2.064877986907959,
"eval_runtime": 53.1644,
"eval_samples_per_second": 4.608,
"eval_steps_per_second": 2.314,
"step": 224
},
{
"epoch": 0.5881499395405079,
"grad_norm": 11.468412399291992,
"learning_rate": 3.749326081899329e-05,
"loss": 2.0694,
"step": 228
},
{
"epoch": 0.5984683595324466,
"grad_norm": 13.881438255310059,
"learning_rate": 3.589529726428615e-05,
"loss": 2.0961,
"step": 232
},
{
"epoch": 0.5984683595324466,
"eval_loss": 2.064319610595703,
"eval_runtime": 31.9203,
"eval_samples_per_second": 7.675,
"eval_steps_per_second": 3.853,
"step": 232
},
{
"epoch": 0.6087867795243853,
"grad_norm": 10.702425956726074,
"learning_rate": 3.431283848421347e-05,
"loss": 2.0564,
"step": 236
},
{
"epoch": 0.619105199516324,
"grad_norm": 11.869012832641602,
"learning_rate": 3.274762401680124e-05,
"loss": 2.0808,
"step": 240
},
{
"epoch": 0.619105199516324,
"eval_loss": 2.056699752807617,
"eval_runtime": 30.4915,
"eval_samples_per_second": 8.035,
"eval_steps_per_second": 4.034,
"step": 240
},
{
"epoch": 0.6294236195082628,
"grad_norm": 9.28243637084961,
"learning_rate": 3.120137444404442e-05,
"loss": 2.1001,
"step": 244
},
{
"epoch": 0.6397420395002016,
"grad_norm": 10.947897911071777,
"learning_rate": 2.9675789500535328e-05,
"loss": 2.1337,
"step": 248
},
{
"epoch": 0.6397420395002016,
"eval_loss": 2.0556507110595703,
"eval_runtime": 31.1947,
"eval_samples_per_second": 7.854,
"eval_steps_per_second": 3.943,
"step": 248
},
{
"epoch": 0.6500604594921403,
"grad_norm": 11.127779006958008,
"learning_rate": 2.8172546205008683e-05,
"loss": 2.0562,
"step": 252
},
{
"epoch": 0.660378879484079,
"grad_norm": 10.932923316955566,
"learning_rate": 2.6693297016857188e-05,
"loss": 2.0565,
"step": 256
},
{
"epoch": 0.660378879484079,
"eval_loss": 2.055548906326294,
"eval_runtime": 31.9066,
"eval_samples_per_second": 7.679,
"eval_steps_per_second": 3.855,
"step": 256
},
{
"epoch": 0.6706972994760177,
"grad_norm": 11.094362258911133,
"learning_rate": 2.523966801964468e-05,
"loss": 2.0685,
"step": 260
},
{
"epoch": 0.6810157194679565,
"grad_norm": 10.753888130187988,
"learning_rate": 2.3813257133612827e-05,
"loss": 2.1184,
"step": 264
},
{
"epoch": 0.6810157194679565,
"eval_loss": 2.0496644973754883,
"eval_runtime": 54.0908,
"eval_samples_per_second": 4.529,
"eval_steps_per_second": 2.274,
"step": 264
},
{
"epoch": 0.6913341394598952,
"grad_norm": 9.6819429397583,
"learning_rate": 2.2415632359146856e-05,
"loss": 2.0437,
"step": 268
},
{
"epoch": 0.7016525594518339,
"grad_norm": 10.608392715454102,
"learning_rate": 2.104833005313131e-05,
"loss": 2.0604,
"step": 272
},
{
"epoch": 0.7016525594518339,
"eval_loss": 2.041249990463257,
"eval_runtime": 34.03,
"eval_samples_per_second": 7.2,
"eval_steps_per_second": 3.614,
"step": 272
},
{
"epoch": 0.7119709794437726,
"grad_norm": 11.31783390045166,
"learning_rate": 1.971285324008994e-05,
"loss": 2.093,
"step": 276
},
{
"epoch": 0.7222893994357114,
"grad_norm": 11.558327674865723,
"learning_rate": 1.84106699599668e-05,
"loss": 2.1099,
"step": 280
},
{
"epoch": 0.7222893994357114,
"eval_loss": 2.038382053375244,
"eval_runtime": 30.4289,
"eval_samples_per_second": 8.052,
"eval_steps_per_second": 4.042,
"step": 280
},
{
"epoch": 0.7326078194276502,
"grad_norm": 9.70297908782959,
"learning_rate": 1.7143211654364762e-05,
"loss": 2.108,
"step": 284
},
{
"epoch": 0.7429262394195889,
"grad_norm": 12.303277969360352,
"learning_rate": 1.5911871593014837e-05,
"loss": 2.1048,
"step": 288
},
{
"epoch": 0.7429262394195889,
"eval_loss": 2.041520118713379,
"eval_runtime": 32.0147,
"eval_samples_per_second": 7.653,
"eval_steps_per_second": 3.842,
"step": 288
},
{
"epoch": 0.7532446594115276,
"grad_norm": 11.557732582092285,
"learning_rate": 1.4718003342206722e-05,
"loss": 2.1051,
"step": 292
},
{
"epoch": 0.7635630794034663,
"grad_norm": 11.619261741638184,
"learning_rate": 1.3562919276863844e-05,
"loss": 2.0692,
"step": 296
},
{
"epoch": 0.7635630794034663,
"eval_loss": 2.033982992172241,
"eval_runtime": 32.7706,
"eval_samples_per_second": 7.476,
"eval_steps_per_second": 3.753,
"step": 296
},
{
"epoch": 0.7738814993954051,
"grad_norm": 10.316109657287598,
"learning_rate": 1.2447889137898293e-05,
"loss": 2.0314,
"step": 300
},
{
"epoch": 0.7841999193873438,
"grad_norm": 12.697169303894043,
"learning_rate": 1.1374138636432053e-05,
"loss": 2.0489,
"step": 304
},
{
"epoch": 0.7841999193873438,
"eval_loss": 2.033088445663452,
"eval_runtime": 50.91,
"eval_samples_per_second": 4.812,
"eval_steps_per_second": 2.416,
"step": 304
},
{
"epoch": 0.7945183393792825,
"grad_norm": 11.112667083740234,
"learning_rate": 1.0342848106418368e-05,
"loss": 2.0759,
"step": 308
},
{
"epoch": 0.8048367593712212,
"grad_norm": 11.560709953308105,
"learning_rate": 9.35515120714447e-06,
"loss": 2.057,
"step": 312
},
{
"epoch": 0.8048367593712212,
"eval_loss": 2.027519702911377,
"eval_runtime": 32.5351,
"eval_samples_per_second": 7.53,
"eval_steps_per_second": 3.781,
"step": 312
},
{
"epoch": 0.8151551793631601,
"grad_norm": 10.368136405944824,
"learning_rate": 8.41213367704224e-06,
"loss": 2.0332,
"step": 316
},
{
"epoch": 0.8254735993550988,
"grad_norm": 9.943156242370605,
"learning_rate": 7.51483214017637e-06,
"loss": 2.0485,
"step": 320
},
{
"epoch": 0.8254735993550988,
"eval_loss": 2.022447347640991,
"eval_runtime": 32.7692,
"eval_samples_per_second": 7.477,
"eval_steps_per_second": 3.754,
"step": 320
},
{
"epoch": 0.8357920193470375,
"grad_norm": 9.394889831542969,
"learning_rate": 6.664232966721995e-06,
"loss": 2.1139,
"step": 324
},
{
"epoch": 0.8461104393389762,
"grad_norm": 8.868988037109375,
"learning_rate": 5.8612711886848196e-06,
"loss": 2.0364,
"step": 328
},
{
"epoch": 0.8461104393389762,
"eval_loss": 2.0201539993286133,
"eval_runtime": 32.5539,
"eval_samples_per_second": 7.526,
"eval_steps_per_second": 3.778,
"step": 328
},
{
"epoch": 0.856428859330915,
"grad_norm": 10.170036315917969,
"learning_rate": 5.106829472055202e-06,
"loss": 2.0615,
"step": 332
},
{
"epoch": 0.8667472793228537,
"grad_norm": 9.667701721191406,
"learning_rate": 4.401737146526219e-06,
"loss": 2.014,
"step": 336
},
{
"epoch": 0.8667472793228537,
"eval_loss": 2.02396821975708,
"eval_runtime": 51.0734,
"eval_samples_per_second": 4.797,
"eval_steps_per_second": 2.408,
"step": 336
},
{
"epoch": 0.8770656993147924,
"grad_norm": 10.14772891998291,
"learning_rate": 3.7467692938425057e-06,
"loss": 2.0966,
"step": 340
},
{
"epoch": 0.8873841193067311,
"grad_norm": 10.002238273620605,
"learning_rate": 3.142645895781715e-06,
"loss": 2.0656,
"step": 344
},
{
"epoch": 0.8873841193067311,
"eval_loss": 2.0236427783966064,
"eval_runtime": 31.6007,
"eval_samples_per_second": 7.753,
"eval_steps_per_second": 3.892,
"step": 344
},
{
"epoch": 0.8977025392986699,
"grad_norm": 8.875214576721191,
"learning_rate": 2.5900310427053044e-06,
"loss": 2.0104,
"step": 348
},
{
"epoch": 0.9080209592906087,
"grad_norm": 9.890630722045898,
"learning_rate": 2.089532203548794e-06,
"loss": 2.0473,
"step": 352
},
{
"epoch": 0.9080209592906087,
"eval_loss": 2.0197205543518066,
"eval_runtime": 33.634,
"eval_samples_per_second": 7.284,
"eval_steps_per_second": 3.657,
"step": 352
},
{
"epoch": 0.9183393792825474,
"grad_norm": 10.363245964050293,
"learning_rate": 1.6416995580537664e-06,
"loss": 2.0972,
"step": 356
},
{
"epoch": 0.9286577992744861,
"grad_norm": 9.002208709716797,
"learning_rate": 1.247025391975698e-06,
"loss": 2.0279,
"step": 360
},
{
"epoch": 0.9286577992744861,
"eval_loss": 2.0180423259735107,
"eval_runtime": 33.0195,
"eval_samples_per_second": 7.42,
"eval_steps_per_second": 3.725,
"step": 360
},
{
"epoch": 0.9389762192664248,
"grad_norm": 9.901611328125,
"learning_rate": 9.059435559326257e-07,
"loss": 2.0544,
"step": 364
},
{
"epoch": 0.9492946392583635,
"grad_norm": 9.303104400634766,
"learning_rate": 6.188289884893062e-07,
"loss": 2.0415,
"step": 368
},
{
"epoch": 0.9492946392583635,
"eval_loss": 2.017828941345215,
"eval_runtime": 31.5389,
"eval_samples_per_second": 7.768,
"eval_steps_per_second": 3.9,
"step": 368
},
{
"epoch": 0.9596130592503023,
"grad_norm": 9.742131233215332,
"learning_rate": 3.8599730400115107e-07,
"loss": 2.0714,
"step": 372
},
{
"epoch": 0.969931479242241,
"grad_norm": 11.851016998291016,
"learning_rate": 2.0770444567118075e-07,
"loss": 2.0419,
"step": 376
},
{
"epoch": 0.969931479242241,
"eval_loss": 2.017824411392212,
"eval_runtime": 51.264,
"eval_samples_per_second": 4.779,
"eval_steps_per_second": 2.399,
"step": 376
}
],
"logging_steps": 4,
"max_steps": 387,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 8,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 5.714203219250381e+17,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}