| [2026-05-11 17:22:50,392][model][WARNING] - Model open-unlearning/tofu_Llama-3.2-1B-Instruct_full requested with {'attn_implementation': 'flash_attention_2'} |
| [2026-05-11 17:22:50,392][model][WARNING] - Model open-unlearning/tofu_Llama-3.2-1B-Instruct_full requested with {'attn_implementation': 'flash_attention_2'} |
| [2026-05-11 17:44:05,751][model][INFO] - Setting pad_token as eos token: <|eot_id|> |
| [2026-05-11 17:44:06,084][model][INFO] - Setting pad_token as eos token: <|eot_id|> |
| [2026-05-11 17:44:11,925][evaluator][INFO] - Evaluations stored in the experiment directory: ./saves/unlearn/tofu_Llama-3.2-1B-Instruct_forget10_GradDiff |
| [2026-05-11 17:44:11,985][evaluator][INFO] - Evaluations stored in the experiment directory: ./saves/unlearn/tofu_Llama-3.2-1B-Instruct_forget10_GradDiff |
| [2026-05-11 17:44:12,957][trainer][INFO] - GradDiff Trainer loaded, output_dir: ./saves/unlearn/tofu_Llama-3.2-1B-Instruct_forget10_GradDiff |
| [2026-05-11 17:44:12,971][trainer][INFO] - GradDiff Trainer loaded, output_dir: ./saves/unlearn/tofu_Llama-3.2-1B-Instruct_forget10_GradDiff |
| [2026-05-11 19:23:21,939][model][INFO] - Setting pad_token as eos token: <|eot_id|> |
| [2026-05-11 19:23:22,984][model][INFO] - Setting pad_token as eos token: <|eot_id|> |
| [2026-05-11 19:23:28,909][evaluator][INFO] - Evaluations stored in the experiment directory: ./saves/unlearn/tofu_Llama-3.2-1B-Instruct_forget10_GradDiff |
| [2026-05-11 19:23:29,838][trainer][INFO] - GradDiff Trainer loaded, output_dir: ./saves/unlearn/tofu_Llama-3.2-1B-Instruct_forget10_GradDiff |
| [2026-05-11 19:30:02,166][model][INFO] - Setting pad_token as eos token: <|eot_id|> |
| [2026-05-11 19:30:09,661][evaluator][INFO] - Evaluations stored in the experiment directory: ./saves/unlearn/tofu_Llama-3.2-1B-Instruct_forget10_GradDiff |
| [2026-05-11 19:30:10,390][trainer][INFO] - GradDiff Trainer loaded, output_dir: ./saves/unlearn/tofu_Llama-3.2-1B-Instruct_forget10_GradDiff |
| [2026-05-11 19:30:15,483][evaluator][INFO] - ***** Running TOFU evaluation suite ***** |
| [2026-05-11 19:30:15,484][evaluator][INFO] - Fine-grained evaluations will be saved to: ./saves/unlearn/tofu_Llama-3.2-1B-Instruct_forget10_GradDiff/checkpoint-0/evals/TOFU_EVAL.json |
| [2026-05-11 19:30:15,484][evaluator][INFO] - Aggregated evaluations will be summarised in: ./saves/unlearn/tofu_Llama-3.2-1B-Instruct_forget10_GradDiff/checkpoint-0/evals/TOFU_SUMMARY.json |
| [2026-05-11 19:30:19,426][metrics][INFO] - Evaluating forget_Q_A_PARA_Prob |
|
|
