Skip to content

support Ascend NPU backend #37

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

Open
MaskerPRC opened this issue May 8, 2025 · 3 comments
Open

support Ascend NPU backend #37

MaskerPRC opened this issue May 8, 2025 · 3 comments

Comments

@MaskerPRC
Copy link

Purpose: Support 3rd-party backend.

Ussage:

"torch.cuda" should be replaced by "device_module"
".cuda()" should be replaced by ".to(device=device_type)"

@MaskerPRC
Copy link
Author

@garrett4wade Does there currently exist a plan to support non-GPU platforms? If there is an opportunity, would you be willing to accept support for abstract device PR?

@garrett4wade
Copy link
Collaborator

Hi, we indeed have a plan to support devices other than NV GPU. Such a PR will be great!

FYI In our code we use torch.cuda for some GPU-related logging. The device type of computation has been written to realhf.base.constants.current_device().

@MaskerPRC
Copy link
Author

Thank you very much, then I am ready to start coding😀

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