| { |
| "best_metric": 0.6603000000000001, |
| "best_model_checkpoint": "text_gen/unlearn/checkpoint/t5-large/neggrad/4.0/samsum_42/checkpoint-219", |
| "epoch": 5.0, |
| "eval_steps": 500, |
| "global_step": 365, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "df_rougeL": 35.3727, |
| "epoch": 1.0, |
| "eval_gen_len": 19.8515625, |
| "eval_loss": 1.6698297262191772, |
| "eval_rouge1": 42.7849, |
| "eval_rouge2": 20.7407, |
| "eval_rougeL": 35.446, |
| "eval_rougeLsum": 39.589, |
| "eval_runtime": 349.4427, |
| "eval_samples_per_second": 2.341, |
| "eval_steps_per_second": 0.037, |
| "eval_unlearn_overall_rougeL": 0.5366499999999981, |
| "step": 73, |
| "test_rougeL": 35.446, |
| "unlearn_overall_rougeL": 0.5366499999999981, |
| "unlearn_time": -1 |
| }, |
| { |
| "df_rougeL": 12.4648, |
| "epoch": 2.0, |
| "eval_gen_len": 9.197916666666666, |
| "eval_loss": 7.590928554534912, |
| "eval_rouge1": 13.9304, |
| "eval_rouge2": 3.6865, |
| "eval_rougeL": 12.3627, |
| "eval_rougeLsum": 13.2181, |
| "eval_runtime": 638.2368, |
| "eval_samples_per_second": 1.282, |
| "eval_steps_per_second": 0.02, |
| "eval_unlearn_overall_rougeL": 0.44894999999999996, |
| "step": 146, |
| "test_rougeL": 12.3627, |
| "unlearn_overall_rougeL": 0.44894999999999996, |
| "unlearn_time": -1 |
| }, |
| { |
| "df_rougeL": 2.9036, |
| "epoch": 3.0, |
| "eval_gen_len": 2.96484375, |
| "eval_loss": 87.1946029663086, |
| "eval_rouge1": 3.236, |
| "eval_rouge2": 0.0575, |
| "eval_rougeL": 3.2242, |
| "eval_rougeLsum": 3.229, |
| "eval_runtime": 65.884, |
| "eval_samples_per_second": 12.416, |
| "eval_steps_per_second": 0.197, |
| "eval_unlearn_overall_rougeL": 0.6603000000000001, |
| "step": 219, |
| "test_rougeL": 3.2242, |
| "unlearn_overall_rougeL": 0.6603000000000001, |
| "unlearn_time": -1 |
| }, |
| { |
| "df_rougeL": 0.9739, |
| "epoch": 4.0, |
| "eval_gen_len": 2.12109375, |
| "eval_loss": 144.74517822265625, |
| "eval_rouge1": 0.8156, |
| "eval_rouge2": 0.0074, |
| "eval_rougeL": 0.8256, |
| "eval_rougeLsum": 0.8157, |
| "eval_runtime": 58.5107, |
| "eval_samples_per_second": 13.98, |
| "eval_steps_per_second": 0.222, |
| "eval_unlearn_overall_rougeL": 0.42585, |
| "step": 292, |
| "test_rougeL": 0.8256, |
| "unlearn_overall_rougeL": 0.42585, |
| "unlearn_time": -1 |
| }, |
| { |
| "df_rougeL": 0.6594, |
| "epoch": 5.0, |
| "eval_gen_len": 2.0377604166666665, |
| "eval_loss": 160.100341796875, |
| "eval_rouge1": 0.6105, |
| "eval_rouge2": 0.0, |
| "eval_rougeL": 0.6166, |
| "eval_rougeLsum": 0.6071, |
| "eval_runtime": 57.912, |
| "eval_samples_per_second": 14.125, |
| "eval_steps_per_second": 0.224, |
| "eval_unlearn_overall_rougeL": 0.4786, |
| "step": 365, |
| "test_rougeL": 0.6166, |
| "unlearn_overall_rougeL": 0.4786, |
| "unlearn_time": -1 |
| }, |
| { |
| "epoch": 5.0, |
| "step": 365, |
| "total_flos": 4443124453244928.0, |
| "train_loss": -37.45592626284247, |
| "train_runtime": 2431.0267, |
| "train_samples_per_second": 1.211, |
| "train_steps_per_second": 0.15 |
| } |
| ], |
| "logging_steps": 1000, |
| "max_steps": 365, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 5, |
| "save_steps": 500, |
| "total_flos": 4443124453244928.0, |
| "train_batch_size": 8, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
