Hyperparameter --lr_finetuning_init and --lr_unlearning_init seems to be not used in ./defense/abl/abl.py

Question

Hyperparameter --lr_finetuning_init and --lr_unlearning_init seems to be not used in ./defense/abl/abl.py

Closed this issue a year ago · 2 comments

Thank you for your excellent work. I found that the hyperparameters --lr_finetuning_init and --lr_unlearning_init seem to be not used in the ./defense/abl/abl.py. Is that a bug?

Answer 1 · 2023-04-14T15:20:24.000Z

Yes, because the experiment found that unlearn had a high sensitivity to lr, so the lr given in the original text was retained without using hyperparameters. In order not to affect you, we will make changes in later versions.

Answer 2 · 2023-04-27T07:59:13.000Z

Ok, thank you for your replies!