| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.011470520761642579, | |
| "eval_steps": 5, | |
| "global_step": 50, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0011470520761642578, | |
| "eval_loss": 5.131626605987549, | |
| "eval_runtime": 13.4035, | |
| "eval_samples_per_second": 491.214, | |
| "eval_steps_per_second": 3.88, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0022941041523285156, | |
| "eval_loss": 5.129306316375732, | |
| "eval_runtime": 13.1975, | |
| "eval_samples_per_second": 498.881, | |
| "eval_steps_per_second": 3.94, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0034411562284927736, | |
| "eval_loss": 5.125307559967041, | |
| "eval_runtime": 13.3602, | |
| "eval_samples_per_second": 492.806, | |
| "eval_steps_per_second": 3.892, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.004588208304657031, | |
| "eval_loss": 5.1195783615112305, | |
| "eval_runtime": 13.3732, | |
| "eval_samples_per_second": 492.33, | |
| "eval_steps_per_second": 3.888, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.005735260380821289, | |
| "eval_loss": 5.11201286315918, | |
| "eval_runtime": 13.3238, | |
| "eval_samples_per_second": 494.153, | |
| "eval_steps_per_second": 3.903, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.006882312456985547, | |
| "eval_loss": 5.102468967437744, | |
| "eval_runtime": 13.3391, | |
| "eval_samples_per_second": 493.586, | |
| "eval_steps_per_second": 3.898, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.008029364533149804, | |
| "eval_loss": 5.090775012969971, | |
| "eval_runtime": 15.3552, | |
| "eval_samples_per_second": 428.779, | |
| "eval_steps_per_second": 3.386, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.009176416609314062, | |
| "eval_loss": 5.0768303871154785, | |
| "eval_runtime": 13.4109, | |
| "eval_samples_per_second": 490.945, | |
| "eval_steps_per_second": 3.877, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01032346868547832, | |
| "eval_loss": 5.0602593421936035, | |
| "eval_runtime": 13.4714, | |
| "eval_samples_per_second": 488.739, | |
| "eval_steps_per_second": 3.86, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.011470520761642579, | |
| "eval_loss": 5.04086446762085, | |
| "eval_runtime": 13.4864, | |
| "eval_samples_per_second": 488.195, | |
| "eval_steps_per_second": 3.856, | |
| "step": 50 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 13077, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 5, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |