Skip to content
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

Fix IA3 config for Falcon models #1007

Merged
merged 2 commits into from
Nov 9, 2023

Conversation

SumanthRH
Copy link
Contributor

What does this PR do?

Minor changes to the default IA 3 config for Falcon models. IA 3 learned vectors are added to the second feedforward layer in the MLP block, and for falcon that would be dense_4h_to_h as seen here.

@HuggingFaceDocBuilderDev

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint.

Copy link
Contributor

@pacman100 pacman100 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hello, thank you @SumanthRH for adding this. However, you would need to add the same to TRANSFORMERS_MODELS_TO_IA3_TARGET_MODULES_MAPPING which is missing.

@SumanthRH
Copy link
Contributor Author

Oops I missed a commit. Added!

@SumanthRH
Copy link
Contributor Author

ping @pacman100

Copy link
Contributor

@pacman100 pacman100 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you @SumanthRH for the fixes to IA3 config mappings, LGTM! 🤗

@pacman100 pacman100 merged commit face67d into huggingface:main Nov 9, 2023
11 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

3 participants