| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 3288, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4562043795620438, | |
| "grad_norm": 6.375082015991211, | |
| "learning_rate": 4.239659367396594e-05, | |
| "loss": 3.2188, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9124087591240876, | |
| "grad_norm": 8.150468826293945, | |
| "learning_rate": 3.479318734793188e-05, | |
| "loss": 3.0547, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.24, | |
| "eval_loss": 2.9638171195983887, | |
| "eval_runtime": 29.1632, | |
| "eval_samples_per_second": 33.433, | |
| "eval_steps_per_second": 4.183, | |
| "step": 1096 | |
| }, | |
| { | |
| "epoch": 1.3686131386861313, | |
| "grad_norm": 9.649592399597168, | |
| "learning_rate": 2.7189781021897807e-05, | |
| "loss": 2.9215, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.8248175182481752, | |
| "grad_norm": 12.843892097473145, | |
| "learning_rate": 1.9586374695863748e-05, | |
| "loss": 2.8673, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.26974358974358975, | |
| "eval_loss": 2.895164966583252, | |
| "eval_runtime": 29.8577, | |
| "eval_samples_per_second": 32.655, | |
| "eval_steps_per_second": 4.086, | |
| "step": 2192 | |
| }, | |
| { | |
| "epoch": 2.281021897810219, | |
| "grad_norm": 9.697604179382324, | |
| "learning_rate": 1.1982968369829685e-05, | |
| "loss": 2.7658, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.7372262773722627, | |
| "grad_norm": 10.829789161682129, | |
| "learning_rate": 4.379562043795621e-06, | |
| "loss": 2.6645, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 3288, | |
| "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": 6924619549799424.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |