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

Turning support? #51

Open
Ph0rk0z opened this issue Nov 27, 2024 · 2 comments
Open

Turning support? #51

Ph0rk0z opened this issue Nov 27, 2024 · 2 comments

Comments

@Ph0rk0z
Copy link

Ph0rk0z commented Nov 27, 2024

Turning has 22 and 24gb cards available as well as being used in google colab. Flash attention skipped this generation causing issues for the cards. Is it possible to add support?

@jt-zhang
Copy link
Member

jt-zhang commented Dec 2, 2024

Hi, we think Turning is something other than a popular architecture. Could you please first try the Triton-only branch and provide some feedback? Thank you.

@Ph0rk0z
Copy link
Author

Ph0rk0z commented Dec 3, 2024

Turning has more affordable 48gb cards too, like RTX 8000. A lot of projects exclude it and require ampere+, hence the popularity is less. I will give triton only a shot. Is the difference that no kernel is used?

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

No branches or pull requests

2 participants