CarD-T / checkpoint-3204 /trainer_state.json
joneill
Added model files and documentation
876243b
{
"best_metric": 0.07832363992929459,
"best_model_checkpoint": "/storage/BioM-ELECTRA-Large-SQuAD2_carDB_5e_neg_lg_SGD/checkpoint-3204",
"epoch": 2.0,
"eval_steps": 500,
"global_step": 3204,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.3121098626716604,
"grad_norm": 3.2046010494232178,
"learning_rate": 4.690387016229713e-05,
"loss": 0.1789,
"step": 500
},
{
"epoch": 0.6242197253433208,
"grad_norm": 1.161939263343811,
"learning_rate": 4.378277153558052e-05,
"loss": 0.1113,
"step": 1000
},
{
"epoch": 0.9363295880149812,
"grad_norm": 2.1341910362243652,
"learning_rate": 4.066167290886392e-05,
"loss": 0.1054,
"step": 1500
},
{
"epoch": 1.0,
"eval_accuracy_score": 0.9690997325938397,
"eval_f1": 0.7794316644113668,
"eval_loss": 0.08875501155853271,
"eval_precision": 0.7747886241352806,
"eval_recall": 0.7841306884480747,
"eval_runtime": 4.093,
"eval_samples_per_second": 348.153,
"eval_steps_per_second": 43.733,
"step": 1602
},
{
"epoch": 1.2484394506866416,
"grad_norm": 0.97569340467453,
"learning_rate": 3.754057428214732e-05,
"loss": 0.079,
"step": 2000
},
{
"epoch": 1.5605493133583022,
"grad_norm": 0.8073092699050903,
"learning_rate": 3.441947565543071e-05,
"loss": 0.0685,
"step": 2500
},
{
"epoch": 1.8726591760299627,
"grad_norm": 2.6909968852996826,
"learning_rate": 3.129837702871411e-05,
"loss": 0.0671,
"step": 3000
},
{
"epoch": 2.0,
"eval_accuracy_score": 0.973506982271962,
"eval_f1": 0.814260389994091,
"eval_loss": 0.07832363992929459,
"eval_precision": 0.8248204309656824,
"eval_recall": 0.8039673278879813,
"eval_runtime": 4.1937,
"eval_samples_per_second": 339.796,
"eval_steps_per_second": 42.683,
"step": 3204
}
],
"logging_steps": 500,
"max_steps": 8010,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"total_flos": 4245597826238592.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}