| { | |
| "best_metric": 0.2796624004840851, | |
| "best_model_checkpoint": "xlm-roberta-large-all-finetuned-toxicity-classification/checkpoint-1126", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 2815, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.644760213143872e-05, | |
| "loss": 0.5369, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.604564905166626, | |
| "eval_runtime": 12.0965, | |
| "eval_samples_per_second": 41.334, | |
| "eval_steps_per_second": 5.208, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 1.2895204262877443e-05, | |
| "loss": 0.3908, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.2796624004840851, | |
| "eval_runtime": 12.1105, | |
| "eval_samples_per_second": 41.287, | |
| "eval_steps_per_second": 5.202, | |
| "step": 1126 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "learning_rate": 9.342806394316164e-06, | |
| "loss": 0.3196, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.2894600033760071, | |
| "eval_runtime": 12.1036, | |
| "eval_samples_per_second": 41.31, | |
| "eval_steps_per_second": 5.205, | |
| "step": 1689 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "learning_rate": 5.790408525754885e-06, | |
| "loss": 0.2412, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.3928073048591614, | |
| "eval_runtime": 12.0987, | |
| "eval_samples_per_second": 41.327, | |
| "eval_steps_per_second": 5.207, | |
| "step": 2252 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "learning_rate": 2.238010657193606e-06, | |
| "loss": 0.1839, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.39615875482559204, | |
| "eval_runtime": 12.0954, | |
| "eval_samples_per_second": 41.338, | |
| "eval_steps_per_second": 5.209, | |
| "step": 2815 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2815, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 2.096845567488e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |