vmpsergio's picture
Training in progress, step 10, checkpoint
4dea035 verified
raw
history blame
2.12 kB
{
"best_metric": 2.0631089210510254,
"best_model_checkpoint": "miner_id_24/checkpoint-10",
"epoch": 0.007136485280999108,
"eval_steps": 5,
"global_step": 10,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0007136485280999108,
"eval_loss": 2.0849006175994873,
"eval_runtime": 166.9589,
"eval_samples_per_second": 3.534,
"eval_steps_per_second": 1.767,
"step": 1
},
{
"epoch": 0.0021409455842997323,
"grad_norm": 0.6842705607414246,
"learning_rate": 6e-05,
"loss": 8.1855,
"step": 3
},
{
"epoch": 0.003568242640499554,
"eval_loss": 2.080864429473877,
"eval_runtime": 168.3342,
"eval_samples_per_second": 3.505,
"eval_steps_per_second": 1.752,
"step": 5
},
{
"epoch": 0.004281891168599465,
"grad_norm": 5.52883243560791,
"learning_rate": 0.00012,
"loss": 8.1743,
"step": 6
},
{
"epoch": 0.006422836752899197,
"grad_norm": 32.8885498046875,
"learning_rate": 0.00018,
"loss": 8.4995,
"step": 9
},
{
"epoch": 0.007136485280999108,
"eval_loss": 2.0631089210510254,
"eval_runtime": 168.2486,
"eval_samples_per_second": 3.507,
"eval_steps_per_second": 1.753,
"step": 10
}
],
"logging_steps": 3,
"max_steps": 30,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 10,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 1,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 0
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 5309400235376640.0,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}