-
Couldn't load subscription status.
- Fork 104
feat: add WMDP dataset integration #93
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
Add WMDP cyber dataset configuration for unlearning
|
|
||
| ### 📢 Updates | ||
|
|
||
| #### [Apr 9, 2025] |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
need to update just before merge
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It's the same as WMDP_cyber_forget.yaml
Is it needed for something? If for testing performance disruption, IIRC MMLU was used.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@ruidazeng I don't see the path existing for this dataset in hf.
cais/wmdp-corpora
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
You mean for MMLU? Yeah, it's not in cais/wmdp-corpora but in https://huggingface.co/datasets/cais/mmlu
From what I understand, in original setup, training is done on:
- "cais/wmdp-corpora" - "cyber-forget-corpus"
- "cais/wmdp-corpora" - "cyber-retain-corpus"
And evaluation on:
- "cais/wmdp" - "wmdp-cyber"
- "cais/mmlu" - "all"
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I see you mentioned in #80 that evaluation won't be done here anyway, but only afterwards, so I guess WMDP_cyber_forget.yaml and WMDP_cyber_retain.yaml just aren't needed.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes
|
Hey @ruidazeng thanks for initiating this work. We now integrated WMDP and lm-eval-harness in #104 . |
Fantastic love the progress! |
Add WMDP cyber dataset configuration for unlearning
What does this PR do?
Fixes #80
Before submitting