microsoft_deberta-base_squad / trainer_state.json
Palak Goenka
add files
51961f3
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 22140,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.07,
"learning_rate": 2.932249322493225e-05,
"loss": 1.6277,
"step": 500
},
{
"epoch": 0.14,
"learning_rate": 2.86449864498645e-05,
"loss": 1.0907,
"step": 1000
},
{
"epoch": 0.2,
"learning_rate": 2.7967479674796748e-05,
"loss": 1.0199,
"step": 1500
},
{
"epoch": 0.27,
"learning_rate": 2.7289972899729e-05,
"loss": 0.9284,
"step": 2000
},
{
"epoch": 0.34,
"learning_rate": 2.6612466124661247e-05,
"loss": 0.9689,
"step": 2500
},
{
"epoch": 0.41,
"learning_rate": 2.59349593495935e-05,
"loss": 0.9188,
"step": 3000
},
{
"epoch": 0.47,
"learning_rate": 2.5257452574525747e-05,
"loss": 0.898,
"step": 3500
},
{
"epoch": 0.54,
"learning_rate": 2.4579945799457998e-05,
"loss": 0.902,
"step": 4000
},
{
"epoch": 0.61,
"learning_rate": 2.3902439024390246e-05,
"loss": 0.8888,
"step": 4500
},
{
"epoch": 0.68,
"learning_rate": 2.3224932249322494e-05,
"loss": 0.8703,
"step": 5000
},
{
"epoch": 0.75,
"learning_rate": 2.2547425474254745e-05,
"loss": 0.842,
"step": 5500
},
{
"epoch": 0.81,
"learning_rate": 2.186991869918699e-05,
"loss": 0.8403,
"step": 6000
},
{
"epoch": 0.88,
"learning_rate": 2.119241192411924e-05,
"loss": 0.8354,
"step": 6500
},
{
"epoch": 0.95,
"learning_rate": 2.051490514905149e-05,
"loss": 0.816,
"step": 7000
},
{
"epoch": 1.02,
"learning_rate": 1.983739837398374e-05,
"loss": 0.7617,
"step": 7500
},
{
"epoch": 1.08,
"learning_rate": 1.915989159891599e-05,
"loss": 0.5977,
"step": 8000
},
{
"epoch": 1.15,
"learning_rate": 1.8482384823848237e-05,
"loss": 0.6169,
"step": 8500
},
{
"epoch": 1.22,
"learning_rate": 1.7804878048780488e-05,
"loss": 0.594,
"step": 9000
},
{
"epoch": 1.29,
"learning_rate": 1.7127371273712736e-05,
"loss": 0.6107,
"step": 9500
},
{
"epoch": 1.36,
"learning_rate": 1.6449864498644987e-05,
"loss": 0.611,
"step": 10000
},
{
"epoch": 1.42,
"learning_rate": 1.5772357723577235e-05,
"loss": 0.63,
"step": 10500
},
{
"epoch": 1.49,
"learning_rate": 1.5094850948509487e-05,
"loss": 0.6039,
"step": 11000
},
{
"epoch": 1.56,
"learning_rate": 1.4417344173441735e-05,
"loss": 0.6143,
"step": 11500
},
{
"epoch": 1.63,
"learning_rate": 1.3739837398373984e-05,
"loss": 0.5886,
"step": 12000
},
{
"epoch": 1.69,
"learning_rate": 1.3062330623306234e-05,
"loss": 0.6002,
"step": 12500
},
{
"epoch": 1.76,
"learning_rate": 1.2384823848238483e-05,
"loss": 0.6013,
"step": 13000
},
{
"epoch": 1.83,
"learning_rate": 1.1707317073170733e-05,
"loss": 0.5779,
"step": 13500
},
{
"epoch": 1.9,
"learning_rate": 1.1029810298102981e-05,
"loss": 0.58,
"step": 14000
},
{
"epoch": 1.96,
"learning_rate": 1.035230352303523e-05,
"loss": 0.5889,
"step": 14500
},
{
"epoch": 2.03,
"learning_rate": 9.674796747967479e-06,
"loss": 0.5174,
"step": 15000
},
{
"epoch": 2.1,
"learning_rate": 8.997289972899728e-06,
"loss": 0.4231,
"step": 15500
},
{
"epoch": 2.17,
"learning_rate": 8.319783197831978e-06,
"loss": 0.4261,
"step": 16000
},
{
"epoch": 2.24,
"learning_rate": 7.642276422764228e-06,
"loss": 0.4206,
"step": 16500
},
{
"epoch": 2.3,
"learning_rate": 6.964769647696477e-06,
"loss": 0.393,
"step": 17000
},
{
"epoch": 2.37,
"learning_rate": 6.287262872628727e-06,
"loss": 0.4052,
"step": 17500
},
{
"epoch": 2.44,
"learning_rate": 5.609756097560976e-06,
"loss": 0.4186,
"step": 18000
},
{
"epoch": 2.51,
"learning_rate": 4.932249322493225e-06,
"loss": 0.4225,
"step": 18500
},
{
"epoch": 2.57,
"learning_rate": 4.254742547425474e-06,
"loss": 0.4012,
"step": 19000
},
{
"epoch": 2.64,
"learning_rate": 3.5772357723577236e-06,
"loss": 0.4124,
"step": 19500
},
{
"epoch": 2.71,
"learning_rate": 2.8997289972899733e-06,
"loss": 0.3895,
"step": 20000
},
{
"epoch": 2.78,
"learning_rate": 2.222222222222222e-06,
"loss": 0.3905,
"step": 20500
},
{
"epoch": 2.85,
"learning_rate": 1.5447154471544717e-06,
"loss": 0.3781,
"step": 21000
},
{
"epoch": 2.91,
"learning_rate": 8.672086720867209e-07,
"loss": 0.3939,
"step": 21500
},
{
"epoch": 2.98,
"learning_rate": 1.8970189701897021e-07,
"loss": 0.408,
"step": 22000
},
{
"epoch": 3.0,
"step": 22140,
"total_flos": 6.07242628548864e+16,
"train_loss": 0.6538377606771826,
"train_runtime": 11699.9166,
"train_samples_per_second": 22.705,
"train_steps_per_second": 1.892
}
],
"max_steps": 22140,
"num_train_epochs": 3,
"total_flos": 6.07242628548864e+16,
"trial_name": null,
"trial_params": null
}