| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.026143790849673, | |
| "eval_steps": 10, | |
| "global_step": 40, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.26143790849673204, | |
| "grad_norm": 80.0328369140625, | |
| "learning_rate": 1.9682539682539684e-05, | |
| "loss": 22.6724, | |
| "mean_token_accuracy": 0.49049999862909316, | |
| "num_tokens": 49120.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.26143790849673204, | |
| "eval_loss": 4.981190204620361, | |
| "eval_mean_token_accuracy": 0.5223076802033645, | |
| "eval_num_tokens": 49120.0, | |
| "eval_runtime": 17.5192, | |
| "eval_samples_per_second": 2.854, | |
| "eval_steps_per_second": 0.742, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.5228758169934641, | |
| "grad_norm": 57.1182746887207, | |
| "learning_rate": 1.8624338624338625e-05, | |
| "loss": 17.5658, | |
| "mean_token_accuracy": 0.5591249987483025, | |
| "num_tokens": 98240.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.5228758169934641, | |
| "eval_loss": 3.730682134628296, | |
| "eval_mean_token_accuracy": 0.5742307580434359, | |
| "eval_num_tokens": 98240.0, | |
| "eval_runtime": 17.4511, | |
| "eval_samples_per_second": 2.865, | |
| "eval_steps_per_second": 0.745, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7843137254901961, | |
| "grad_norm": 58.24352264404297, | |
| "learning_rate": 1.7566137566137566e-05, | |
| "loss": 13.113, | |
| "mean_token_accuracy": 0.6388749912381172, | |
| "num_tokens": 147360.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.7843137254901961, | |
| "eval_loss": 2.7529046535491943, | |
| "eval_mean_token_accuracy": 0.7115384431985708, | |
| "eval_num_tokens": 147360.0, | |
| "eval_runtime": 17.6786, | |
| "eval_samples_per_second": 2.828, | |
| "eval_steps_per_second": 0.735, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.026143790849673, | |
| "grad_norm": 30.73020362854004, | |
| "learning_rate": 1.6507936507936507e-05, | |
| "loss": 8.7369, | |
| "mean_token_accuracy": 0.7555405417004147, | |
| "num_tokens": 192796.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.026143790849673, | |
| "eval_loss": 1.836286187171936, | |
| "eval_mean_token_accuracy": 0.8476923199800345, | |
| "eval_num_tokens": 192796.0, | |
| "eval_runtime": 17.4538, | |
| "eval_samples_per_second": 2.865, | |
| "eval_steps_per_second": 0.745, | |
| "step": 40 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 195, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 20, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5013303699161472.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |