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

Graph and fusion optimisation disabled in XPU branch during model training #734

Open
Wetitpig opened this issue Nov 15, 2024 · 1 comment
Open

Graph and fusion optimisation disabled in XPU branch during model training #734

Wetitpig opened this issue Nov 15, 2024 · 1 comment
Assignees
@devpramod

Comments

@Wetitpig
Copy link

Describe the issue

When I am trying to trace a module on XPU, a warning is raised in training:

/**/lib/python3.11/site-packages/torch/nn/modules/module.py:1541: UserWarning: The grad mode is detected as torch.no_grad() is NOT enabled. In this mode on XPU, please expect NO graph and fusion optimization will be applied. 
 (Triggered internally at /build/intel-pytorch-extension/csrc/gpu/jit/fusion_pass.cpp:829.)

May I ask why are operator fusions disabled in grad mode in 568869c?
I don't think it is well-explained anywhere in the commit notes.
Thank you
@devpramod devpramod self-assigned this Nov 26, 2024
@jingxu10
Copy link
Contributor

jingxu10 commented Dec 4, 2024

@gujinghui @tye1 any idea?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@devpramod devpramod

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
3 participants
@Wetitpig @jingxu10 @devpramod