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鈥檒l occasionally send you account related emails.

Already on GitHub? Sign in to your account

Generate: starcoder 馃 馃 assisted generation #23182

Merged
merged 2 commits into from
May 8, 2023

Conversation

gante
Copy link
Member

@gante gante commented May 6, 2023

What does this PR do?

Starcoder (GPTBigCode) has a unique cache format, and assisted generation is heavy on cache-related ops. This PR adds the GPTBigCode special case.

All slow tests for assisted generation are passing after these changes.

@gante gante requested a review from sgugger May 6, 2023 16:06
@HuggingFaceDocBuilderDev
Copy link

HuggingFaceDocBuilderDev commented May 6, 2023

The documentation is not available anymore as the PR was closed or merged.

Copy link
Collaborator

@sgugger sgugger left a comment

Choose a reason for hiding this comment

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

Arg. Would love to be able to change those parts directly in the model files in the future, instead of hacking special keys like this :-/

Thanks for your PR!

@gante
Copy link
Member Author

gante commented May 8, 2023

Arg. Would love to be able to change those parts directly in the model files in the future, instead of hacking special keys like this :-/

@sgugger I'm going to add a common test for the cache format, to ensure we don't do this again for future models :)

@gante gante merged commit bbfb9fc into huggingface:main May 8, 2023
4 checks passed
@gante gante deleted the starcoder_assisted branch May 8, 2023 09:45
gojiteji pushed a commit to gojiteji/transformers that referenced this pull request Jun 5, 2023
* starcoder has joined the chat

* indexing that works for all
novice03 pushed a commit to novice03/transformers that referenced this pull request Jun 23, 2023
* starcoder has joined the chat

* indexing that works for all
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