trained_model / trainer_state.json
peterbeamish's picture
End of training
9ff1453
raw
history blame
2.08 kB
{
"best_metric": 1.3398905992507935,
"best_model_checkpoint": "./trained_model/checkpoint-200",
"epoch": 6.0,
"eval_steps": 100,
"global_step": 300,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 2.0,
"learning_rate": 2.6666666666666667e-05,
"loss": 1.9101,
"step": 100
},
{
"epoch": 2.0,
"eval_gen_len": 12.7,
"eval_loss": 1.4032484292984009,
"eval_rouge1": 0.3928,
"eval_rouge2": 0.2897,
"eval_rougeL": 0.3915,
"eval_rougeLsum": 0.3911,
"eval_runtime": 14.5472,
"eval_samples_per_second": 6.874,
"eval_steps_per_second": 3.437,
"step": 100
},
{
"epoch": 4.0,
"learning_rate": 1.3333333333333333e-05,
"loss": 1.4264,
"step": 200
},
{
"epoch": 4.0,
"eval_gen_len": 12.68,
"eval_loss": 1.3398905992507935,
"eval_rouge1": 0.3826,
"eval_rouge2": 0.2908,
"eval_rougeL": 0.3808,
"eval_rougeLsum": 0.3809,
"eval_runtime": 14.6388,
"eval_samples_per_second": 6.831,
"eval_steps_per_second": 3.416,
"step": 200
},
{
"epoch": 6.0,
"learning_rate": 0.0,
"loss": 1.3121,
"step": 300
},
{
"epoch": 6.0,
"eval_gen_len": 13.16,
"eval_loss": 1.3339253664016724,
"eval_rouge1": 0.3878,
"eval_rouge2": 0.2902,
"eval_rougeL": 0.3871,
"eval_rougeLsum": 0.3871,
"eval_runtime": 15.4855,
"eval_samples_per_second": 6.458,
"eval_steps_per_second": 3.229,
"step": 300
},
{
"epoch": 6.0,
"step": 300,
"total_flos": 111534302822400.0,
"train_loss": 1.5495406595865886,
"train_runtime": 75.3909,
"train_samples_per_second": 7.959,
"train_steps_per_second": 3.979
}
],
"logging_steps": 100,
"max_steps": 300,
"num_train_epochs": 6,
"save_steps": 200,
"total_flos": 111534302822400.0,
"trial_name": null,
"trial_params": null
}