mt_coref_en_ru_gender_exp / trainer_state.json
cattana's picture
init
bdb8824
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 46248,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03,
"learning_rate": 4.945943608372254e-05,
"loss": 1.9245,
"step": 500
},
{
"epoch": 0.06,
"learning_rate": 4.891887216744508e-05,
"loss": 1.8947,
"step": 1000
},
{
"epoch": 0.1,
"learning_rate": 4.837830825116762e-05,
"loss": 1.862,
"step": 1500
},
{
"epoch": 0.13,
"learning_rate": 4.783774433489016e-05,
"loss": 1.8378,
"step": 2000
},
{
"epoch": 0.16,
"learning_rate": 4.72971804186127e-05,
"loss": 1.8137,
"step": 2500
},
{
"epoch": 0.19,
"learning_rate": 4.675661650233524e-05,
"loss": 1.8268,
"step": 3000
},
{
"epoch": 0.23,
"learning_rate": 4.621605258605778e-05,
"loss": 1.833,
"step": 3500
},
{
"epoch": 0.26,
"learning_rate": 4.567548866978032e-05,
"loss": 1.8099,
"step": 4000
},
{
"epoch": 0.29,
"learning_rate": 4.513492475350286e-05,
"loss": 1.797,
"step": 4500
},
{
"epoch": 0.32,
"learning_rate": 4.4594360837225395e-05,
"loss": 1.8208,
"step": 5000
},
{
"epoch": 0.36,
"learning_rate": 4.405379692094794e-05,
"loss": 1.7937,
"step": 5500
},
{
"epoch": 0.39,
"learning_rate": 4.351323300467047e-05,
"loss": 1.7805,
"step": 6000
},
{
"epoch": 0.42,
"learning_rate": 4.2972669088393016e-05,
"loss": 1.7785,
"step": 6500
},
{
"epoch": 0.45,
"learning_rate": 4.2432105172115553e-05,
"loss": 1.7621,
"step": 7000
},
{
"epoch": 0.49,
"learning_rate": 4.189154125583809e-05,
"loss": 1.7649,
"step": 7500
},
{
"epoch": 0.52,
"learning_rate": 4.135097733956063e-05,
"loss": 1.7694,
"step": 8000
},
{
"epoch": 0.55,
"learning_rate": 4.0810413423283174e-05,
"loss": 1.7598,
"step": 8500
},
{
"epoch": 0.58,
"learning_rate": 4.0269849507005705e-05,
"loss": 1.7519,
"step": 9000
},
{
"epoch": 0.62,
"learning_rate": 3.972928559072825e-05,
"loss": 1.766,
"step": 9500
},
{
"epoch": 0.65,
"learning_rate": 3.918872167445079e-05,
"loss": 1.7489,
"step": 10000
},
{
"epoch": 0.68,
"learning_rate": 3.8648157758173326e-05,
"loss": 1.7528,
"step": 10500
},
{
"epoch": 0.71,
"learning_rate": 3.8107593841895864e-05,
"loss": 1.7669,
"step": 11000
},
{
"epoch": 0.75,
"learning_rate": 3.756702992561841e-05,
"loss": 1.7279,
"step": 11500
},
{
"epoch": 0.78,
"learning_rate": 3.7026466009340946e-05,
"loss": 1.727,
"step": 12000
},
{
"epoch": 0.81,
"learning_rate": 3.6485902093063484e-05,
"loss": 1.7368,
"step": 12500
},
{
"epoch": 0.84,
"learning_rate": 3.594533817678602e-05,
"loss": 1.7203,
"step": 13000
},
{
"epoch": 0.88,
"learning_rate": 3.540477426050857e-05,
"loss": 1.7159,
"step": 13500
},
{
"epoch": 0.91,
"learning_rate": 3.4864210344231105e-05,
"loss": 1.7376,
"step": 14000
},
{
"epoch": 0.94,
"learning_rate": 3.432364642795364e-05,
"loss": 1.703,
"step": 14500
},
{
"epoch": 0.97,
"learning_rate": 3.378308251167619e-05,
"loss": 1.7146,
"step": 15000
},
{
"epoch": 1.01,
"learning_rate": 3.324251859539872e-05,
"loss": 1.7062,
"step": 15500
},
{
"epoch": 1.04,
"learning_rate": 3.270195467912126e-05,
"loss": 1.5423,
"step": 16000
},
{
"epoch": 1.07,
"learning_rate": 3.21613907628438e-05,
"loss": 1.5257,
"step": 16500
},
{
"epoch": 1.1,
"learning_rate": 3.162082684656634e-05,
"loss": 1.5494,
"step": 17000
},
{
"epoch": 1.14,
"learning_rate": 3.108026293028888e-05,
"loss": 1.5378,
"step": 17500
},
{
"epoch": 1.17,
"learning_rate": 3.053969901401142e-05,
"loss": 1.5513,
"step": 18000
},
{
"epoch": 1.2,
"learning_rate": 2.999913509773396e-05,
"loss": 1.5425,
"step": 18500
},
{
"epoch": 1.23,
"learning_rate": 2.9458571181456497e-05,
"loss": 1.5442,
"step": 19000
},
{
"epoch": 1.26,
"learning_rate": 2.8918007265179035e-05,
"loss": 1.5424,
"step": 19500
},
{
"epoch": 1.3,
"learning_rate": 2.8377443348901576e-05,
"loss": 1.5471,
"step": 20000
},
{
"epoch": 1.33,
"learning_rate": 2.7836879432624114e-05,
"loss": 1.5515,
"step": 20500
},
{
"epoch": 1.36,
"learning_rate": 2.7296315516346656e-05,
"loss": 1.5548,
"step": 21000
},
{
"epoch": 1.39,
"learning_rate": 2.6755751600069197e-05,
"loss": 1.5423,
"step": 21500
},
{
"epoch": 1.43,
"learning_rate": 2.621518768379173e-05,
"loss": 1.5581,
"step": 22000
},
{
"epoch": 1.46,
"learning_rate": 2.5674623767514273e-05,
"loss": 1.5137,
"step": 22500
},
{
"epoch": 1.49,
"learning_rate": 2.5134059851236814e-05,
"loss": 1.5299,
"step": 23000
},
{
"epoch": 1.52,
"learning_rate": 2.4593495934959352e-05,
"loss": 1.5392,
"step": 23500
},
{
"epoch": 1.56,
"learning_rate": 2.405293201868189e-05,
"loss": 1.5539,
"step": 24000
},
{
"epoch": 1.59,
"learning_rate": 2.351236810240443e-05,
"loss": 1.5464,
"step": 24500
},
{
"epoch": 1.62,
"learning_rate": 2.297180418612697e-05,
"loss": 1.5622,
"step": 25000
},
{
"epoch": 1.65,
"learning_rate": 2.2431240269849507e-05,
"loss": 1.5474,
"step": 25500
},
{
"epoch": 1.69,
"learning_rate": 2.189067635357205e-05,
"loss": 1.5424,
"step": 26000
},
{
"epoch": 1.72,
"learning_rate": 2.1350112437294586e-05,
"loss": 1.5372,
"step": 26500
},
{
"epoch": 1.75,
"learning_rate": 2.0809548521017124e-05,
"loss": 1.537,
"step": 27000
},
{
"epoch": 1.78,
"learning_rate": 2.0268984604739665e-05,
"loss": 1.551,
"step": 27500
},
{
"epoch": 1.82,
"learning_rate": 1.9728420688462203e-05,
"loss": 1.5424,
"step": 28000
},
{
"epoch": 1.85,
"learning_rate": 1.9187856772184745e-05,
"loss": 1.526,
"step": 28500
},
{
"epoch": 1.88,
"learning_rate": 1.8647292855907283e-05,
"loss": 1.5433,
"step": 29000
},
{
"epoch": 1.91,
"learning_rate": 1.810672893962982e-05,
"loss": 1.5287,
"step": 29500
},
{
"epoch": 1.95,
"learning_rate": 1.7566165023352362e-05,
"loss": 1.531,
"step": 30000
},
{
"epoch": 1.98,
"learning_rate": 1.70256011070749e-05,
"loss": 1.5208,
"step": 30500
},
{
"epoch": 2.01,
"learning_rate": 1.648503719079744e-05,
"loss": 1.4804,
"step": 31000
},
{
"epoch": 2.04,
"learning_rate": 1.594447327451998e-05,
"loss": 1.4182,
"step": 31500
},
{
"epoch": 2.08,
"learning_rate": 1.540390935824252e-05,
"loss": 1.3974,
"step": 32000
},
{
"epoch": 2.11,
"learning_rate": 1.486334544196506e-05,
"loss": 1.3985,
"step": 32500
},
{
"epoch": 2.14,
"learning_rate": 1.4322781525687598e-05,
"loss": 1.3941,
"step": 33000
},
{
"epoch": 2.17,
"learning_rate": 1.3782217609410136e-05,
"loss": 1.3968,
"step": 33500
},
{
"epoch": 2.21,
"learning_rate": 1.3241653693132677e-05,
"loss": 1.4081,
"step": 34000
},
{
"epoch": 2.24,
"learning_rate": 1.2701089776855215e-05,
"loss": 1.409,
"step": 34500
},
{
"epoch": 2.27,
"learning_rate": 1.2160525860577754e-05,
"loss": 1.4122,
"step": 35000
},
{
"epoch": 2.3,
"learning_rate": 1.1619961944300296e-05,
"loss": 1.4138,
"step": 35500
},
{
"epoch": 2.34,
"learning_rate": 1.1079398028022835e-05,
"loss": 1.4052,
"step": 36000
},
{
"epoch": 2.37,
"learning_rate": 1.0538834111745373e-05,
"loss": 1.39,
"step": 36500
},
{
"epoch": 2.4,
"learning_rate": 9.998270195467913e-06,
"loss": 1.4037,
"step": 37000
},
{
"epoch": 2.43,
"learning_rate": 9.457706279190453e-06,
"loss": 1.4046,
"step": 37500
},
{
"epoch": 2.46,
"learning_rate": 8.917142362912992e-06,
"loss": 1.398,
"step": 38000
},
{
"epoch": 2.5,
"learning_rate": 8.37657844663553e-06,
"loss": 1.4122,
"step": 38500
},
{
"epoch": 2.53,
"learning_rate": 7.83601453035807e-06,
"loss": 1.3967,
"step": 39000
},
{
"epoch": 2.56,
"learning_rate": 7.295450614080609e-06,
"loss": 1.3963,
"step": 39500
},
{
"epoch": 2.59,
"learning_rate": 6.754886697803149e-06,
"loss": 1.4131,
"step": 40000
},
{
"epoch": 2.63,
"learning_rate": 6.2143227815256884e-06,
"loss": 1.381,
"step": 40500
},
{
"epoch": 2.66,
"learning_rate": 5.673758865248227e-06,
"loss": 1.4123,
"step": 41000
},
{
"epoch": 2.69,
"learning_rate": 5.133194948970767e-06,
"loss": 1.4132,
"step": 41500
},
{
"epoch": 2.72,
"learning_rate": 4.5926310326933056e-06,
"loss": 1.4029,
"step": 42000
},
{
"epoch": 2.76,
"learning_rate": 4.052067116415845e-06,
"loss": 1.3981,
"step": 42500
},
{
"epoch": 2.79,
"learning_rate": 3.5115032001383844e-06,
"loss": 1.3978,
"step": 43000
},
{
"epoch": 2.82,
"learning_rate": 2.970939283860924e-06,
"loss": 1.4081,
"step": 43500
},
{
"epoch": 2.85,
"learning_rate": 2.430375367583463e-06,
"loss": 1.4115,
"step": 44000
},
{
"epoch": 2.89,
"learning_rate": 1.8898114513060028e-06,
"loss": 1.4018,
"step": 44500
},
{
"epoch": 2.92,
"learning_rate": 1.3492475350285417e-06,
"loss": 1.3923,
"step": 45000
},
{
"epoch": 2.95,
"learning_rate": 8.086836187510811e-07,
"loss": 1.3863,
"step": 45500
},
{
"epoch": 2.98,
"learning_rate": 2.681197024736205e-07,
"loss": 1.3951,
"step": 46000
},
{
"epoch": 3.0,
"step": 46248,
"total_flos": 1.7013927281688576e+16,
"train_loss": 1.5740079802226246,
"train_runtime": 16384.5833,
"train_samples_per_second": 45.162,
"train_steps_per_second": 2.823
}
],
"max_steps": 46248,
"num_train_epochs": 3,
"total_flos": 1.7013927281688576e+16,
"trial_name": null,
"trial_params": null
}