| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.680965147453083, | |
| "eval_steps": 500, | |
| "global_step": 2000, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.6702412868632708, | |
| "grad_norm": 2.054403066635132, | |
| "learning_rate": 6.75599008099931e-05, | |
| "loss": 0.6856, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.4924572706222534, | |
| "eval_loss": 0.6931472420692444, | |
| "eval_runtime": 8.9909, | |
| "eval_samples_per_second": 331.779, | |
| "eval_steps_per_second": 20.799, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.3404825737265416, | |
| "grad_norm": 1.88927161693573, | |
| "learning_rate": 6.270645965984992e-05, | |
| "loss": 0.7049, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5129064917564392, | |
| "eval_loss": 0.6931472420692444, | |
| "eval_runtime": 9.0209, | |
| "eval_samples_per_second": 330.678, | |
| "eval_steps_per_second": 20.73, | |
| "step": 1492 | |
| }, | |
| { | |
| "epoch": 2.0107238605898123, | |
| "grad_norm": 1.888950228691101, | |
| "learning_rate": 5.7853018509706724e-05, | |
| "loss": 0.7041, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.680965147453083, | |
| "grad_norm": 1.8515383005142212, | |
| "learning_rate": 5.299957735956355e-05, | |
| "loss": 0.7017, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 7460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 5210880251877900.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 7.241334196013627e-05, | |
| "per_device_train_batch_size": 16 | |
| } | |
| } | |