Traceback (most recent call last):
File "FLAT/unlearn.py", line 2, in <module>
from dataloader import CustomTrainerForgetting, custom_data_collator_forget, custom_data_collator_forget_hp, custom_data_collator_forget_hp_cl, custom_data_collator_forget_hp_idk, custom_data_collator_forget_hp_dpo
File "FLAT/dataloader.py", line 7, in <module>
from evaluate_util import get_dataloader, get_all_evals
ModuleNotFoundError: No module named 'evaluate_util'
It seems the code you provided does not include the evaluate_util.py file.
Also, in dataloader.py, the code for handling the loss_type 'GLS' is missing, which is the default configuration in forget.yaml.
Could you please update the code to make it reproducible?
Additionally, please provide the code for other benchmarks such as TOFU and MUSE if possible.
It seems the code you provided does not include the
evaluate_util.pyfile.Also, in
dataloader.py, the code for handling the loss_type 'GLS' is missing, which is the default configuration inforget.yaml.Could you please update the code to make it reproducible?
Additionally, please provide the code for other benchmarks such as TOFU and MUSE if possible.