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

runtime memory leak #352

Closed
Mddct opened this issue May 1, 2021 · 8 comments
Closed

runtime memory leak #352

Mddct opened this issue May 1, 2021 · 8 comments
Labels
good first issue Good for newcomers

Comments

@Mddct
Copy link
Collaborator

Mddct commented May 1, 2021

libtorch model cause memory growth for courrent call.

@robin1001
Copy link
Collaborator

Is there any solution to fix it?

@Mddct
Copy link
Collaborator Author

Mddct commented May 6, 2021

I tried some environment variables but none of them worked. Maybe need to recompile libtorch without mkl

@lvzhiqiang
Copy link

Has the problem benn solved? @Mddct

@Mddct
Copy link
Collaborator Author

Mddct commented May 17, 2021

@lvzhiqiang
sorry,I've been busy recently, and haven't tried to solve it yet。

@robin1001 robin1001 reopened this Oct 29, 2021
@day9011
Copy link
Contributor

day9011 commented Oct 29, 2021

@lvzhiqiang sorry,I've been busy recently, and haven't tried to solve it yet。

Try LRU_CACHE_CAPACITY=1

@Mddct
Copy link
Collaborator Author

Mddct commented Nov 8, 2021

Is there any solution to fix it?
Maybe related issues:
pytorch/pytorch#27971 (comment)

@xingchensong
Copy link
Member

Will hold this PR for anyone else who has encountered and interested in the same question.

@xingchensong
Copy link
Member

Please note that memory leak has been fixed by jemalloc in PR #1011

@xingchensong xingchensong added good first issue Good for newcomers and removed question Further information is requested labels Feb 21, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
good first issue Good for newcomers
Projects
None yet
Development

No branches or pull requests

5 participants