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.

Sign up for GitHub

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

Jump to bottom

why dense_embedding_codegen_lookup_function has No DispatchKey::Autograd with DataParallel #2657

Open
imhandsome opened this issue Dec 27, 2024 · 6 comments
Open

why dense_embedding_codegen_lookup_function has No DispatchKey::Autograd with DataParallel #2657

imhandsome opened this issue Dec 27, 2024 · 6 comments

Comments

@imhandsome
Copy link

image
@imhandsome
Copy link
Author

It has backward method，But Autograd is not registered

@sarckk
Copy link
Member

sarckk commented Jan 3, 2025

@imhandsome are you training on GPU or CPU?

@imhandsome
Copy link
Author

gpu

@JacoCheung
Copy link

Having the same issue

@imhandsome
Copy link
Author

@sarckk GPU

@sarckk
Copy link
Member

sarckk commented Jan 8, 2025 •
edited
Loading

@imhandsome I think this is just a user warning saying that when using PyTorch 2 (torch.compile) you may silently run into an incorrect behaviour. However if you aren't using PT2 shouldn't be an issue in your case.

I believe a fix would be to update this op's autograd registration, like in pytorch/FBGEMM#1948. @sryap any plans to do that here?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
3 participants
@imhandsome @JacoCheung @sarckk