-
Notifications
You must be signed in to change notification settings - Fork 1k
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
ipex intel extension for pytorch integration #1255
Conversation
Co-authored-by: Sourab Mangrulkar <[email protected]> Co-authored-by: jianan-gu <[email protected]> Co-authored-by: Wang, Yi A <[email protected]>
@sgugger @pacman100 @yao-matrix @jianan-gu please review |
The documentation is not available anymore as the PR was closed or merged. |
Signed-off-by: Wang, Yi A <[email protected]>
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.
Thanks for opening a fresh PR. I have some comments, and we would also need to add tests for this.
I add testcase, please help review, later, I will add doc about this |
Signed-off-by: Wang, Yi A <[email protected]>
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.
Thanks for your work on this! @pacman100 could you have a second look?
@pacman100 could you help with the review? |
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.
Hello, Thank you @sywangyi for adding ipex support and iterating on the previous PR 🤗. I believe adding docs with the below details would help users to get a better idea about when and how to use it:
- Brief intro on what ipex does and what is integrated currently in accelerate (only training for now)
- How to use the current integration
- Link to blog/original docs showing benefits when using ipex on CPU in comparison to without it
For reference, please find the DeepSpeed integration docs here: https://huggingface.co/docs/accelerate/usage_guides/deepspeed
agree, I will create a new PR for the document. For this PR, could you help with the merge? |
What does this PR do?
Implements feature request #700