roberta-base-bne-finetuned-pos / trainer_state.json
josecannete's picture
adding model finetuned on POS
1f9078e
{
"best_metric": 0.032893065363168716,
"best_model_checkpoint": "/home/jcanete/ft-data/all_results/pos/roberta_base_bne/epochs_2_bs_16_lr_5e-5/checkpoint-1600",
"epoch": 2.0,
"global_step": 1790,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.22,
"eval_accuracy": 0.984541433111356,
"eval_f1": 0.9812364013813204,
"eval_loss": 0.06434043496847153,
"eval_precision": 0.9803362383582953,
"eval_recall": 0.982138219016603,
"eval_runtime": 2.9618,
"eval_samples_per_second": 558.44,
"eval_steps_per_second": 35.114,
"step": 200
},
{
"epoch": 0.45,
"eval_accuracy": 0.9877832606716811,
"eval_f1": 0.9851088887558137,
"eval_loss": 0.04959944263100624,
"eval_precision": 0.9842051732046347,
"eval_recall": 0.9860142654492418,
"eval_runtime": 2.9534,
"eval_samples_per_second": 560.023,
"eval_steps_per_second": 35.213,
"step": 400
},
{
"epoch": 0.56,
"learning_rate": 3.603351955307263e-05,
"loss": 0.1416,
"step": 500
},
{
"epoch": 0.67,
"eval_accuracy": 0.9888212842747409,
"eval_f1": 0.9869598210719706,
"eval_loss": 0.0426110178232193,
"eval_precision": 0.9864673359813178,
"eval_recall": 0.9874527981458912,
"eval_runtime": 2.954,
"eval_samples_per_second": 559.912,
"eval_steps_per_second": 35.206,
"step": 600
},
{
"epoch": 0.89,
"eval_accuracy": 0.9898912470655872,
"eval_f1": 0.9885165361878895,
"eval_loss": 0.03891567513346672,
"eval_precision": 0.9881021300381291,
"eval_recall": 0.9889312900841142,
"eval_runtime": 2.9458,
"eval_samples_per_second": 561.485,
"eval_steps_per_second": 35.305,
"step": 800
},
{
"epoch": 1.12,
"learning_rate": 2.206703910614525e-05,
"loss": 0.0348,
"step": 1000
},
{
"epoch": 1.12,
"eval_accuracy": 0.9904182436640636,
"eval_f1": 0.9885959656480927,
"eval_loss": 0.03714817017316818,
"eval_precision": 0.9882209666793108,
"eval_recall": 0.9889712493256878,
"eval_runtime": 2.9497,
"eval_samples_per_second": 560.735,
"eval_steps_per_second": 35.258,
"step": 1000
},
{
"epoch": 1.34,
"eval_accuracy": 0.990514061227423,
"eval_f1": 0.9887465425824039,
"eval_loss": 0.03583846241235733,
"eval_precision": 0.9883024592781858,
"eval_recall": 0.9891910251543425,
"eval_runtime": 2.9981,
"eval_samples_per_second": 551.674,
"eval_steps_per_second": 34.688,
"step": 1200
},
{
"epoch": 1.56,
"eval_accuracy": 0.9909133010747536,
"eval_f1": 0.9891582142001439,
"eval_loss": 0.03380614146590233,
"eval_precision": 0.9885068938683481,
"eval_recall": 0.9898103933987333,
"eval_runtime": 2.9525,
"eval_samples_per_second": 560.202,
"eval_steps_per_second": 35.224,
"step": 1400
},
{
"epoch": 1.68,
"learning_rate": 8.100558659217877e-06,
"loss": 0.0202,
"step": 1500
},
{
"epoch": 1.79,
"eval_accuracy": 0.9914882064549099,
"eval_f1": 0.9899645514004692,
"eval_loss": 0.032893065363168716,
"eval_precision": 0.9895396654289935,
"eval_recall": 0.9903898024015504,
"eval_runtime": 2.9564,
"eval_samples_per_second": 559.471,
"eval_steps_per_second": 35.178,
"step": 1600
},
{
"epoch": 2.0,
"step": 1790,
"total_flos": 1190583063933120.0,
"train_loss": 0.05748698165296842,
"train_runtime": 136.7323,
"train_samples_per_second": 209.241,
"train_steps_per_second": 13.091
}
],
"max_steps": 1790,
"num_train_epochs": 2,
"total_flos": 1190583063933120.0,
"trial_name": null,
"trial_params": null
}