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

Update compression config for psmathur/orca_mini_3b #862

Merged

Conversation

nikita-savelyevv
Copy link
Collaborator

What does this PR do?

Disable AWQ for psmathur/orca_mini_3b model.

Reason for changes

Applying AWQ leads to higher PPL compared to without AWQ:

Precision PPL
FP16 16.95
INT4, without AWQ 17.44
INT4, with AWQ 17.51

Before submitting

  • This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
  • Did you make sure to update the documentation with your changes?
  • Did you write any new necessary tests?

@nikita-savelyevv
Copy link
Collaborator Author

@AlexKoff88 @KodiaqQ please take a look

@AlexKoff88
Copy link
Collaborator

@andreyanufr, looks like a case to analyze

@HuggingFaceDocBuilderDev

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

@AlexKoff88 AlexKoff88 merged commit cb32118 into huggingface:main Aug 10, 2024
17 checks passed
github-merge-queue bot pushed a commit to openvinotoolkit/openvino.genai that referenced this pull request Aug 12, 2024
Since AWQ is disabled and ratio is 1.0, there is no need for statistics
collection.

Related to huggingface/optimum-intel#862
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.

4 participants