| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 48, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 5.076903300918695, | |
| "learning_rate": 5e-06, | |
| "loss": 1.2219, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.09993314743042, | |
| "eval_runtime": 11.5006, | |
| "eval_samples_per_second": 37.302, | |
| "eval_steps_per_second": 0.609, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 2.856173758705409, | |
| "learning_rate": 5e-06, | |
| "loss": 1.09, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.875, | |
| "grad_norm": 2.887739861358555, | |
| "learning_rate": 5e-06, | |
| "loss": 1.031, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.0452321767807007, | |
| "eval_runtime": 11.6521, | |
| "eval_samples_per_second": 36.817, | |
| "eval_steps_per_second": 0.601, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.3574686077424067, | |
| "learning_rate": 5e-06, | |
| "loss": 0.9746, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.021416425704956, | |
| "eval_runtime": 10.5751, | |
| "eval_samples_per_second": 40.567, | |
| "eval_steps_per_second": 0.662, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 48, | |
| "total_flos": 80192408125440.0, | |
| "train_loss": 1.0573215981324513, | |
| "train_runtime": 1884.1504, | |
| "train_samples_per_second": 12.966, | |
| "train_steps_per_second": 0.025 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 48, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 80192408125440.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |