| { |
| "best_global_step": 1252, |
| "best_metric": 0.8706804513931274, |
| "best_model_checkpoint": "classifier-clinc-MBbase-distilled/run-1/checkpoint-1252", |
| "epoch": 4.0, |
| "eval_steps": 500, |
| "global_step": 1252, |
| "is_hyper_param_search": true, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.8841935483870967, |
| "eval_loss": 2.7125542163848877, |
| "eval_runtime": 8.642, |
| "eval_samples_per_second": 358.712, |
| "eval_steps_per_second": 11.224, |
| "step": 313 |
| }, |
| { |
| "epoch": 1.5974440894568689, |
| "grad_norm": 15.764195442199707, |
| "learning_rate": 1.468583599574015e-05, |
| "loss": 5.4057, |
| "step": 500 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.9380645161290323, |
| "eval_loss": 1.3560537099838257, |
| "eval_runtime": 8.8142, |
| "eval_samples_per_second": 351.707, |
| "eval_steps_per_second": 11.005, |
| "step": 626 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.9490322580645161, |
| "eval_loss": 1.0215872526168823, |
| "eval_runtime": 8.8098, |
| "eval_samples_per_second": 351.879, |
| "eval_steps_per_second": 11.01, |
| "step": 939 |
| }, |
| { |
| "epoch": 3.194888178913738, |
| "grad_norm": 3.932048797607422, |
| "learning_rate": 9.361022364217253e-06, |
| "loss": 1.0728, |
| "step": 1000 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_accuracy": 0.9538709677419355, |
| "eval_loss": 0.8706804513931274, |
| "eval_runtime": 8.7191, |
| "eval_samples_per_second": 355.54, |
| "eval_steps_per_second": 11.125, |
| "step": 1252 |
| } |
| ], |
| "logging_steps": 500, |
| "max_steps": 1878, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 6, |
| "save_steps": 500, |
| "stateful_callbacks": { |
| "EarlyStoppingCallback": { |
| "args": { |
| "early_stopping_patience": 2, |
| "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": 445404172075776.0, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": { |
| "alpha": 0.43669134703064955, |
| "num_train_epochs": 6, |
| "temperature": 4.700820204359401 |
| } |
| } |
|
|
