distilbert_km_50_v1_mrpc / trainer_state.json
Hartunka's picture
End of training
fef4a86 verified
{
"best_global_step": 30,
"best_metric": 0.5999014377593994,
"best_model_checkpoint": "distilbert_km_50_v1_mrpc/checkpoint-30",
"epoch": 7.0,
"eval_steps": 500,
"global_step": 105,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 1.0,
"grad_norm": 1.2432717084884644,
"learning_rate": 4.9e-05,
"loss": 0.6261,
"step": 15
},
{
"epoch": 1.0,
"eval_accuracy": 0.6838235294117647,
"eval_combined_score": 0.7416007227152136,
"eval_f1": 0.7993779160186625,
"eval_loss": 0.6057308912277222,
"eval_runtime": 0.1977,
"eval_samples_per_second": 2063.225,
"eval_steps_per_second": 10.114,
"step": 15
},
{
"epoch": 2.0,
"grad_norm": 1.6871485710144043,
"learning_rate": 4.8e-05,
"loss": 0.5774,
"step": 30
},
{
"epoch": 2.0,
"eval_accuracy": 0.6985294117647058,
"eval_combined_score": 0.7559414888869053,
"eval_f1": 0.8133535660091047,
"eval_loss": 0.5999014377593994,
"eval_runtime": 0.2046,
"eval_samples_per_second": 1994.194,
"eval_steps_per_second": 9.775,
"step": 30
},
{
"epoch": 3.0,
"grad_norm": 1.3563528060913086,
"learning_rate": 4.7e-05,
"loss": 0.5242,
"step": 45
},
{
"epoch": 3.0,
"eval_accuracy": 0.6838235294117647,
"eval_combined_score": 0.7437382487241472,
"eval_f1": 0.8036529680365296,
"eval_loss": 0.627457857131958,
"eval_runtime": 0.1819,
"eval_samples_per_second": 2242.955,
"eval_steps_per_second": 10.995,
"step": 45
},
{
"epoch": 4.0,
"grad_norm": 1.6578878164291382,
"learning_rate": 4.600000000000001e-05,
"loss": 0.4621,
"step": 60
},
{
"epoch": 4.0,
"eval_accuracy": 0.6764705882352942,
"eval_combined_score": 0.7228506787330318,
"eval_f1": 0.7692307692307693,
"eval_loss": 0.6474722027778625,
"eval_runtime": 0.1826,
"eval_samples_per_second": 2234.357,
"eval_steps_per_second": 10.953,
"step": 60
},
{
"epoch": 5.0,
"grad_norm": 2.2386577129364014,
"learning_rate": 4.5e-05,
"loss": 0.3665,
"step": 75
},
{
"epoch": 5.0,
"eval_accuracy": 0.6691176470588235,
"eval_combined_score": 0.7233561876150789,
"eval_f1": 0.7775947281713345,
"eval_loss": 0.7383757829666138,
"eval_runtime": 0.2072,
"eval_samples_per_second": 1968.96,
"eval_steps_per_second": 9.652,
"step": 75
},
{
"epoch": 6.0,
"grad_norm": 1.8673553466796875,
"learning_rate": 4.4000000000000006e-05,
"loss": 0.2353,
"step": 90
},
{
"epoch": 6.0,
"eval_accuracy": 0.6397058823529411,
"eval_combined_score": 0.6902233115468409,
"eval_f1": 0.7407407407407407,
"eval_loss": 0.9133227467536926,
"eval_runtime": 0.2057,
"eval_samples_per_second": 1983.545,
"eval_steps_per_second": 9.723,
"step": 90
},
{
"epoch": 7.0,
"grad_norm": 3.354482650756836,
"learning_rate": 4.3e-05,
"loss": 0.1316,
"step": 105
},
{
"epoch": 7.0,
"eval_accuracy": 0.6421568627450981,
"eval_combined_score": 0.6956488781079442,
"eval_f1": 0.7491408934707904,
"eval_loss": 1.1885764598846436,
"eval_runtime": 0.204,
"eval_samples_per_second": 1999.806,
"eval_steps_per_second": 9.803,
"step": 105
},
{
"epoch": 7.0,
"step": 105,
"total_flos": 1700616463945728.0,
"train_loss": 0.4176197801317487,
"train_runtime": 34.4638,
"train_samples_per_second": 5321.527,
"train_steps_per_second": 21.762
}
],
"logging_steps": 1,
"max_steps": 750,
"num_input_tokens_seen": 0,
"num_train_epochs": 50,
"save_steps": 500,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 5,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 5
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1700616463945728.0,
"train_batch_size": 256,
"trial_name": null,
"trial_params": null
}