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

Ranks code update - OOM solution #22

Closed
wants to merge 3 commits into from

Conversation

SwapnilDreams100
Copy link
Contributor

Seems like .long() is causing a bump in the memory since it creates a new tensor from the bool, and shouldn't be necessary.
Also I deleted an unused variable to free up memory.
Hopefully should solve the OOM issues.

@joerunde
Copy link
Collaborator

joerunde commented May 3, 2024

(Not to be merged into this repo since this is core vllm code)

@SwapnilDreams100
Copy link
Contributor Author

Hi Joe, Actually I'm not a 100% sure if it solves the issue yet, since we didnt capture any of the failure test cases.
So was thinking we could try it out on BAM first to see if any other failure cases occur(hence PR'd here), and if its resolved, then push upstream. What do you think?

Co-authored-by: Nick Hill <nickhill@us.ibm.com>
Signed-off-by: Swapnil Parekh <swapnilbp100@gmail.com>
@njhill
Copy link
Member

njhill commented May 13, 2024

Thanks again @SwapnilDreams100! Closing this now that it's been merged upstream.

@njhill njhill closed this May 13, 2024
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

Successfully merging this pull request may close these issues.

3 participants