| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 4611, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.32530904359141183, | |
| "grad_norm": 0.7525109052658081, | |
| "learning_rate": 4.4578182606809806e-05, | |
| "loss": 0.2922, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6506180871828237, | |
| "grad_norm": 1.5843052864074707, | |
| "learning_rate": 3.91563652136196e-05, | |
| "loss": 0.1019, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.9759271307742355, | |
| "grad_norm": 1.2410870790481567, | |
| "learning_rate": 3.3734547820429406e-05, | |
| "loss": 0.0743, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.3012361743656473, | |
| "grad_norm": 1.6869269609451294, | |
| "learning_rate": 2.8312730427239213e-05, | |
| "loss": 0.0566, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.6265452179570592, | |
| "grad_norm": 0.7089857459068298, | |
| "learning_rate": 2.2890913034049014e-05, | |
| "loss": 0.0514, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.951854261548471, | |
| "grad_norm": 1.388731837272644, | |
| "learning_rate": 1.7469095640858817e-05, | |
| "loss": 0.05, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.277163305139883, | |
| "grad_norm": 2.429743528366089, | |
| "learning_rate": 1.204727824766862e-05, | |
| "loss": 0.0374, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.6024723487312946, | |
| "grad_norm": 3.7513585090637207, | |
| "learning_rate": 6.625460854478421e-06, | |
| "loss": 0.0346, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.9277813923227067, | |
| "grad_norm": 0.6153927445411682, | |
| "learning_rate": 1.203643461288224e-06, | |
| "loss": 0.0363, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 4611, | |
| "total_flos": 71142932377296.0, | |
| "train_loss": 0.08043386847769664, | |
| "train_runtime": 787.9023, | |
| "train_samples_per_second": 46.795, | |
| "train_steps_per_second": 5.852 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4611, | |
| "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": 71142932377296.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |