chinese_extract_longbert / trainer_state.json
frett's picture
End of training
9311223 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 4320,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.5787037037037037,
"grad_norm": 8.216769218444824,
"learning_rate": 2.652777777777778e-05,
"loss": 4.5636,
"step": 500
},
{
"epoch": 1.1574074074074074,
"grad_norm": 11.628066062927246,
"learning_rate": 2.3055555555555554e-05,
"loss": 3.763,
"step": 1000
},
{
"epoch": 1.7361111111111112,
"grad_norm": 9.473577499389648,
"learning_rate": 1.9583333333333333e-05,
"loss": 3.4828,
"step": 1500
},
{
"epoch": 2.314814814814815,
"grad_norm": 10.071398735046387,
"learning_rate": 1.6111111111111115e-05,
"loss": 3.308,
"step": 2000
},
{
"epoch": 2.8935185185185186,
"grad_norm": 10.017462730407715,
"learning_rate": 1.2638888888888888e-05,
"loss": 3.1879,
"step": 2500
},
{
"epoch": 3.4722222222222223,
"grad_norm": 12.130489349365234,
"learning_rate": 9.166666666666668e-06,
"loss": 3.0,
"step": 3000
},
{
"epoch": 4.050925925925926,
"grad_norm": 13.29631233215332,
"learning_rate": 5.694444444444445e-06,
"loss": 2.9502,
"step": 3500
},
{
"epoch": 4.62962962962963,
"grad_norm": 10.901876449584961,
"learning_rate": 2.222222222222222e-06,
"loss": 2.8179,
"step": 4000
},
{
"epoch": 5.0,
"step": 4320,
"total_flos": 3.611513123226624e+16,
"train_loss": 3.3395631154378256,
"train_runtime": 2548.4435,
"train_samples_per_second": 54.235,
"train_steps_per_second": 1.695
}
],
"logging_steps": 500,
"max_steps": 4320,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.611513123226624e+16,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}