| { |
| "best_metric": 0.77145, |
| "best_model_checkpoint": "text_gen/unlearn/checkpoint/t5-large/bad_teaching/2.0/samsum_42/checkpoint-18", |
| "epoch": 5.0, |
| "eval_steps": 500, |
| "global_step": 90, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "df_rougeL": 2.5684, |
| "epoch": 1.0, |
| "eval_gen_len": 117.5125, |
| "eval_loss": 5.078139781951904, |
| "eval_rouge1": 3.4658, |
| "eval_rouge2": 1.3784, |
| "eval_rougeL": 3.1345, |
| "eval_rougeLsum": 3.2705, |
| "eval_runtime": 1496.5741, |
| "eval_samples_per_second": 0.547, |
| "eval_steps_per_second": 0.017, |
| "eval_unlearn_overall_rougeL": 0.77145, |
| "step": 18, |
| "test_rougeL": 3.1113, |
| "unlearn_overall_rougeL": 0.77145, |
| "unlearn_time": -1 |
| }, |
| { |
| "df_rougeL": 16.3877, |
| "epoch": 2.0, |
| "eval_gen_len": 9.90375, |
| "eval_loss": 2.7741811275482178, |
| "eval_rouge1": 18.0247, |
| "eval_rouge2": 8.3532, |
| "eval_rougeL": 16.3652, |
| "eval_rougeLsum": 17.022, |
| "eval_runtime": 263.4378, |
| "eval_samples_per_second": 3.105, |
| "eval_steps_per_second": 0.099, |
| "eval_unlearn_overall_rougeL": 0.4784500000000005, |
| "step": 36, |
| "test_rougeL": 16.3446, |
| "unlearn_overall_rougeL": 0.4784500000000005, |
| "unlearn_time": -1 |
| }, |
| { |
| "df_rougeL": 0.037, |
| "epoch": 3.0, |
| "eval_gen_len": 120.74375, |
| "eval_loss": 32.48212432861328, |
| "eval_rouge1": 0.1096, |
| "eval_rouge2": 0.0147, |
| "eval_rougeL": 0.1078, |
| "eval_rougeLsum": 0.1103, |
| "eval_runtime": 1499.9815, |
| "eval_samples_per_second": 0.545, |
| "eval_steps_per_second": 0.017, |
| "eval_unlearn_overall_rougeL": 0.5387, |
| "step": 54, |
| "test_rougeL": 0.1144, |
| "unlearn_overall_rougeL": 0.5387, |
| "unlearn_time": -1 |
| }, |
| { |
| "df_rougeL": 0.037, |
| "epoch": 4.0, |
| "eval_gen_len": 125.785, |
| "eval_loss": 38.382774353027344, |
| "eval_rouge1": 0.0702, |
| "eval_rouge2": 0.0179, |
| "eval_rougeL": 0.0691, |
| "eval_rougeLsum": 0.0696, |
| "eval_runtime": 1500.0591, |
| "eval_samples_per_second": 0.545, |
| "eval_steps_per_second": 0.017, |
| "eval_unlearn_overall_rougeL": 0.5177499999999999, |
| "step": 72, |
| "test_rougeL": 0.0725, |
| "unlearn_overall_rougeL": 0.5177499999999999, |
| "unlearn_time": -1 |
| }, |
| { |
| "df_rougeL": 0.037, |
| "epoch": 5.0, |
| "eval_gen_len": 126.54, |
| "eval_loss": 39.759925842285156, |
| "eval_rouge1": 0.0426, |
| "eval_rouge2": 0.0, |
| "eval_rougeL": 0.0423, |
| "eval_rougeLsum": 0.0426, |
| "eval_runtime": 1500.0105, |
| "eval_samples_per_second": 0.545, |
| "eval_steps_per_second": 0.017, |
| "eval_unlearn_overall_rougeL": 0.50285, |
| "step": 90, |
| "test_rougeL": 0.0427, |
| "unlearn_overall_rougeL": 0.50285, |
| "unlearn_time": -1 |
| }, |
| { |
| "epoch": 5.0, |
| "step": 90, |
| "total_flos": 5937370337378304.0, |
| "train_loss": 0.0, |
| "train_runtime": 8983.6984, |
| "train_samples_per_second": 0.327, |
| "train_steps_per_second": 0.01 |
| } |
| ], |
| "logging_steps": 1000, |
| "max_steps": 90, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 5, |
| "save_steps": 500, |
| "total_flos": 5937370337378304.0, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
