Davlan's picture
Upload trainer_state.json
0fae682
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9999768663104076,
"global_step": 64839,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"learning_rate": 4.9614429587131204e-05,
"loss": 1.9292,
"step": 500
},
{
"epoch": 0.05,
"learning_rate": 4.9228859174262406e-05,
"loss": 1.8232,
"step": 1000
},
{
"epoch": 0.07,
"learning_rate": 4.884328876139361e-05,
"loss": 1.7474,
"step": 1500
},
{
"epoch": 0.09,
"learning_rate": 4.845771834852481e-05,
"loss": 1.6787,
"step": 2000
},
{
"epoch": 0.12,
"learning_rate": 4.807214793565602e-05,
"loss": 1.6332,
"step": 2500
},
{
"epoch": 0.14,
"learning_rate": 4.768657752278721e-05,
"loss": 1.584,
"step": 3000
},
{
"epoch": 0.16,
"learning_rate": 4.7301007109918414e-05,
"loss": 1.5654,
"step": 3500
},
{
"epoch": 0.19,
"learning_rate": 4.691543669704962e-05,
"loss": 1.5113,
"step": 4000
},
{
"epoch": 0.21,
"learning_rate": 4.652986628418082e-05,
"loss": 1.4909,
"step": 4500
},
{
"epoch": 0.23,
"learning_rate": 4.614429587131202e-05,
"loss": 1.4654,
"step": 5000
},
{
"epoch": 0.25,
"learning_rate": 4.575872545844323e-05,
"loss": 1.4466,
"step": 5500
},
{
"epoch": 0.28,
"learning_rate": 4.537315504557442e-05,
"loss": 1.4305,
"step": 6000
},
{
"epoch": 0.3,
"learning_rate": 4.4987584632705625e-05,
"loss": 1.3963,
"step": 6500
},
{
"epoch": 0.32,
"learning_rate": 4.4602014219836826e-05,
"loss": 1.388,
"step": 7000
},
{
"epoch": 0.35,
"learning_rate": 4.4216443806968035e-05,
"loss": 1.3647,
"step": 7500
},
{
"epoch": 0.37,
"learning_rate": 4.383087339409923e-05,
"loss": 1.3549,
"step": 8000
},
{
"epoch": 0.39,
"learning_rate": 4.344530298123043e-05,
"loss": 1.336,
"step": 8500
},
{
"epoch": 0.42,
"learning_rate": 4.305973256836164e-05,
"loss": 1.3306,
"step": 9000
},
{
"epoch": 0.44,
"learning_rate": 4.2674162155492835e-05,
"loss": 1.3242,
"step": 9500
},
{
"epoch": 0.46,
"learning_rate": 4.2288591742624036e-05,
"loss": 1.3018,
"step": 10000
},
{
"epoch": 0.49,
"learning_rate": 4.1903021329755245e-05,
"loss": 1.2957,
"step": 10500
},
{
"epoch": 0.51,
"learning_rate": 4.151745091688644e-05,
"loss": 1.2851,
"step": 11000
},
{
"epoch": 0.53,
"learning_rate": 4.113188050401764e-05,
"loss": 1.2768,
"step": 11500
},
{
"epoch": 0.56,
"learning_rate": 4.074631009114885e-05,
"loss": 1.2619,
"step": 12000
},
{
"epoch": 0.58,
"learning_rate": 4.036073967828005e-05,
"loss": 1.2513,
"step": 12500
},
{
"epoch": 0.6,
"learning_rate": 3.997516926541125e-05,
"loss": 1.2438,
"step": 13000
},
{
"epoch": 0.62,
"learning_rate": 3.9589598852542455e-05,
"loss": 1.2194,
"step": 13500
},
{
"epoch": 0.65,
"learning_rate": 3.920402843967366e-05,
"loss": 1.2121,
"step": 14000
},
{
"epoch": 0.67,
"learning_rate": 3.881845802680485e-05,
"loss": 1.2242,
"step": 14500
},
{
"epoch": 0.69,
"learning_rate": 3.843288761393606e-05,
"loss": 1.2131,
"step": 15000
},
{
"epoch": 0.72,
"learning_rate": 3.804731720106726e-05,
"loss": 1.2157,
"step": 15500
},
{
"epoch": 0.74,
"learning_rate": 3.7661746788198464e-05,
"loss": 1.1866,
"step": 16000
},
{
"epoch": 0.76,
"learning_rate": 3.7276176375329665e-05,
"loss": 1.1897,
"step": 16500
},
{
"epoch": 0.79,
"learning_rate": 3.689060596246087e-05,
"loss": 1.181,
"step": 17000
},
{
"epoch": 0.81,
"learning_rate": 3.650503554959207e-05,
"loss": 1.1739,
"step": 17500
},
{
"epoch": 0.83,
"learning_rate": 3.611946513672327e-05,
"loss": 1.1757,
"step": 18000
},
{
"epoch": 0.86,
"learning_rate": 3.573389472385447e-05,
"loss": 1.1519,
"step": 18500
},
{
"epoch": 0.88,
"learning_rate": 3.5348324310985674e-05,
"loss": 1.1544,
"step": 19000
},
{
"epoch": 0.9,
"learning_rate": 3.4962753898116876e-05,
"loss": 1.1543,
"step": 19500
},
{
"epoch": 0.93,
"learning_rate": 3.457718348524808e-05,
"loss": 1.1473,
"step": 20000
},
{
"epoch": 0.95,
"learning_rate": 3.419161307237928e-05,
"loss": 1.1385,
"step": 20500
},
{
"epoch": 0.97,
"learning_rate": 3.380604265951048e-05,
"loss": 1.1219,
"step": 21000
},
{
"epoch": 0.99,
"learning_rate": 3.342047224664168e-05,
"loss": 1.1335,
"step": 21500
},
{
"epoch": 1.02,
"learning_rate": 3.3034901833772884e-05,
"loss": 1.1096,
"step": 22000
},
{
"epoch": 1.04,
"learning_rate": 3.2649331420904086e-05,
"loss": 1.1093,
"step": 22500
},
{
"epoch": 1.06,
"learning_rate": 3.226376100803529e-05,
"loss": 1.1033,
"step": 23000
},
{
"epoch": 1.09,
"learning_rate": 3.187819059516649e-05,
"loss": 1.0932,
"step": 23500
},
{
"epoch": 1.11,
"learning_rate": 3.149262018229769e-05,
"loss": 1.0927,
"step": 24000
},
{
"epoch": 1.13,
"learning_rate": 3.110704976942889e-05,
"loss": 1.0957,
"step": 24500
},
{
"epoch": 1.16,
"learning_rate": 3.0721479356560094e-05,
"loss": 1.0855,
"step": 25000
},
{
"epoch": 1.18,
"learning_rate": 3.03359089436913e-05,
"loss": 1.089,
"step": 25500
},
{
"epoch": 1.2,
"learning_rate": 2.99503385308225e-05,
"loss": 1.0788,
"step": 26000
},
{
"epoch": 1.23,
"learning_rate": 2.95647681179537e-05,
"loss": 1.0798,
"step": 26500
},
{
"epoch": 1.25,
"learning_rate": 2.9179197705084905e-05,
"loss": 1.07,
"step": 27000
},
{
"epoch": 1.27,
"learning_rate": 2.8793627292216106e-05,
"loss": 1.0736,
"step": 27500
},
{
"epoch": 1.3,
"learning_rate": 2.8408056879347305e-05,
"loss": 1.0503,
"step": 28000
},
{
"epoch": 1.32,
"learning_rate": 2.802248646647851e-05,
"loss": 1.0581,
"step": 28500
},
{
"epoch": 1.34,
"learning_rate": 2.763691605360971e-05,
"loss": 1.048,
"step": 29000
},
{
"epoch": 1.36,
"learning_rate": 2.725134564074091e-05,
"loss": 1.044,
"step": 29500
},
{
"epoch": 1.39,
"learning_rate": 2.6865775227872115e-05,
"loss": 1.0392,
"step": 30000
},
{
"epoch": 1.41,
"learning_rate": 2.6480204815003317e-05,
"loss": 1.0381,
"step": 30500
},
{
"epoch": 1.43,
"learning_rate": 2.609463440213452e-05,
"loss": 1.0404,
"step": 31000
},
{
"epoch": 1.46,
"learning_rate": 2.570906398926572e-05,
"loss": 1.0296,
"step": 31500
},
{
"epoch": 1.48,
"learning_rate": 2.532349357639692e-05,
"loss": 1.0272,
"step": 32000
},
{
"epoch": 1.5,
"learning_rate": 2.4937923163528123e-05,
"loss": 1.0349,
"step": 32500
},
{
"epoch": 1.53,
"learning_rate": 2.455235275065933e-05,
"loss": 1.0093,
"step": 33000
},
{
"epoch": 1.55,
"learning_rate": 2.4166782337790527e-05,
"loss": 1.0055,
"step": 33500
},
{
"epoch": 1.57,
"learning_rate": 2.3781211924921732e-05,
"loss": 1.0172,
"step": 34000
},
{
"epoch": 1.6,
"learning_rate": 2.3395641512052934e-05,
"loss": 1.007,
"step": 34500
},
{
"epoch": 1.62,
"learning_rate": 2.3010071099184132e-05,
"loss": 0.9954,
"step": 35000
},
{
"epoch": 1.64,
"learning_rate": 2.2624500686315337e-05,
"loss": 1.0058,
"step": 35500
},
{
"epoch": 1.67,
"learning_rate": 2.223893027344654e-05,
"loss": 1.0001,
"step": 36000
},
{
"epoch": 1.69,
"learning_rate": 2.185335986057774e-05,
"loss": 0.9934,
"step": 36500
},
{
"epoch": 1.71,
"learning_rate": 2.1467789447708942e-05,
"loss": 0.9829,
"step": 37000
},
{
"epoch": 1.74,
"learning_rate": 2.1082219034840144e-05,
"loss": 0.9994,
"step": 37500
},
{
"epoch": 1.76,
"learning_rate": 2.0696648621971345e-05,
"loss": 0.9782,
"step": 38000
},
{
"epoch": 1.78,
"learning_rate": 2.0311078209102547e-05,
"loss": 0.9843,
"step": 38500
},
{
"epoch": 1.8,
"learning_rate": 1.992550779623375e-05,
"loss": 0.981,
"step": 39000
},
{
"epoch": 1.83,
"learning_rate": 1.953993738336495e-05,
"loss": 0.98,
"step": 39500
},
{
"epoch": 1.85,
"learning_rate": 1.9154366970496156e-05,
"loss": 0.9663,
"step": 40000
},
{
"epoch": 1.87,
"learning_rate": 1.8768796557627354e-05,
"loss": 0.9705,
"step": 40500
},
{
"epoch": 1.9,
"learning_rate": 1.8383226144758556e-05,
"loss": 0.9871,
"step": 41000
},
{
"epoch": 1.92,
"learning_rate": 1.799765573188976e-05,
"loss": 0.9679,
"step": 41500
},
{
"epoch": 1.94,
"learning_rate": 1.761208531902096e-05,
"loss": 0.9672,
"step": 42000
},
{
"epoch": 1.97,
"learning_rate": 1.7226514906152164e-05,
"loss": 0.9618,
"step": 42500
},
{
"epoch": 1.99,
"learning_rate": 1.6840944493283366e-05,
"loss": 0.9621,
"step": 43000
},
{
"epoch": 2.01,
"learning_rate": 1.6455374080414564e-05,
"loss": 0.9445,
"step": 43500
},
{
"epoch": 2.04,
"learning_rate": 1.606980366754577e-05,
"loss": 0.9448,
"step": 44000
},
{
"epoch": 2.06,
"learning_rate": 1.568423325467697e-05,
"loss": 0.9405,
"step": 44500
},
{
"epoch": 2.08,
"learning_rate": 1.5298662841808173e-05,
"loss": 0.9352,
"step": 45000
},
{
"epoch": 2.11,
"learning_rate": 1.4913092428939374e-05,
"loss": 0.9308,
"step": 45500
},
{
"epoch": 2.13,
"learning_rate": 1.4527522016070574e-05,
"loss": 0.9281,
"step": 46000
},
{
"epoch": 2.15,
"learning_rate": 1.4141951603201778e-05,
"loss": 0.9344,
"step": 46500
},
{
"epoch": 2.17,
"learning_rate": 1.3756381190332978e-05,
"loss": 0.9337,
"step": 47000
},
{
"epoch": 2.2,
"learning_rate": 1.3370810777464183e-05,
"loss": 0.9272,
"step": 47500
},
{
"epoch": 2.22,
"learning_rate": 1.2985240364595383e-05,
"loss": 0.9302,
"step": 48000
},
{
"epoch": 2.24,
"learning_rate": 1.2599669951726583e-05,
"loss": 0.9297,
"step": 48500
},
{
"epoch": 2.27,
"learning_rate": 1.2214099538857788e-05,
"loss": 0.9238,
"step": 49000
},
{
"epoch": 2.29,
"learning_rate": 1.1828529125988988e-05,
"loss": 0.9181,
"step": 49500
},
{
"epoch": 2.31,
"learning_rate": 1.144295871312019e-05,
"loss": 0.9133,
"step": 50000
},
{
"epoch": 2.34,
"learning_rate": 1.1057388300251391e-05,
"loss": 0.9244,
"step": 50500
},
{
"epoch": 2.36,
"learning_rate": 1.0671817887382595e-05,
"loss": 0.9086,
"step": 51000
},
{
"epoch": 2.38,
"learning_rate": 1.0286247474513797e-05,
"loss": 0.9055,
"step": 51500
},
{
"epoch": 2.41,
"learning_rate": 9.900677061644998e-06,
"loss": 0.9084,
"step": 52000
},
{
"epoch": 2.43,
"learning_rate": 9.5151066487762e-06,
"loss": 0.9082,
"step": 52500
},
{
"epoch": 2.45,
"learning_rate": 9.129536235907402e-06,
"loss": 0.9062,
"step": 53000
},
{
"epoch": 2.48,
"learning_rate": 8.743965823038603e-06,
"loss": 0.8988,
"step": 53500
},
{
"epoch": 2.5,
"learning_rate": 8.358395410169805e-06,
"loss": 0.9025,
"step": 54000
},
{
"epoch": 2.52,
"learning_rate": 7.972824997301008e-06,
"loss": 0.8846,
"step": 54500
},
{
"epoch": 2.54,
"learning_rate": 7.5872545844322085e-06,
"loss": 0.9002,
"step": 55000
},
{
"epoch": 2.57,
"learning_rate": 7.201684171563411e-06,
"loss": 0.8875,
"step": 55500
},
{
"epoch": 2.59,
"learning_rate": 6.816113758694613e-06,
"loss": 0.8902,
"step": 56000
},
{
"epoch": 2.61,
"learning_rate": 6.430543345825815e-06,
"loss": 0.894,
"step": 56500
},
{
"epoch": 2.64,
"learning_rate": 6.044972932957017e-06,
"loss": 0.8845,
"step": 57000
},
{
"epoch": 2.66,
"learning_rate": 5.659402520088219e-06,
"loss": 0.8866,
"step": 57500
},
{
"epoch": 2.68,
"learning_rate": 5.273832107219421e-06,
"loss": 0.8833,
"step": 58000
},
{
"epoch": 2.71,
"learning_rate": 4.888261694350622e-06,
"loss": 0.8906,
"step": 58500
},
{
"epoch": 2.73,
"learning_rate": 4.502691281481825e-06,
"loss": 0.8869,
"step": 59000
},
{
"epoch": 2.75,
"learning_rate": 4.117120868613026e-06,
"loss": 0.8848,
"step": 59500
},
{
"epoch": 2.78,
"learning_rate": 3.731550455744228e-06,
"loss": 0.8744,
"step": 60000
},
{
"epoch": 2.8,
"learning_rate": 3.34598004287543e-06,
"loss": 0.8698,
"step": 60500
},
{
"epoch": 2.82,
"learning_rate": 2.960409630006632e-06,
"loss": 0.876,
"step": 61000
},
{
"epoch": 2.85,
"learning_rate": 2.574839217137834e-06,
"loss": 0.8739,
"step": 61500
},
{
"epoch": 2.87,
"learning_rate": 2.1892688042690357e-06,
"loss": 0.8827,
"step": 62000
},
{
"epoch": 2.89,
"learning_rate": 1.8036983914002379e-06,
"loss": 0.8773,
"step": 62500
},
{
"epoch": 2.91,
"learning_rate": 1.4181279785314396e-06,
"loss": 0.8599,
"step": 63000
},
{
"epoch": 2.94,
"learning_rate": 1.0325575656626413e-06,
"loss": 0.8711,
"step": 63500
},
{
"epoch": 2.96,
"learning_rate": 6.469871527938432e-07,
"loss": 0.8778,
"step": 64000
},
{
"epoch": 2.98,
"learning_rate": 2.614167399250451e-07,
"loss": 0.8713,
"step": 64500
},
{
"epoch": 3.0,
"step": 64839,
"total_flos": 6.047604206657864e+17,
"train_loss": 1.089792190311953,
"train_runtime": 32594.2653,
"train_samples_per_second": 19.893,
"train_steps_per_second": 1.989
}
],
"max_steps": 64839,
"num_train_epochs": 3,
"total_flos": 6.047604206657864e+17,
"trial_name": null,
"trial_params": null
}