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

Custom CUDA extensions #137

Closed
4 of 6 tasks
msaroufim opened this issue Apr 15, 2024 · 0 comments
Closed
4 of 6 tasks

Custom CUDA extensions #137

msaroufim opened this issue Apr 15, 2024 · 0 comments

Comments

@msaroufim
Copy link
Member

msaroufim commented Apr 15, 2024

We'd like to make it really easy for people to add support for custom CUDA extensions in ao and there's a few pieces of work we need to do to get there

Follow up work in a separate issue

  • Make an example without premium runners - you can build cuda extensions without a cuda machine per @malfet
  • Add a useful kernel people should be using like paged attention
yanbing-j pushed a commit to yanbing-j/ao that referenced this issue Dec 9, 2024
* dtype test

* quantized ops fixes

* undo linear8 op

* rename dtype test

* tab->spc

* dtype fix

* disable int4 until pytorch/pytorch#123794 lands

* dtype
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

1 participant