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

Fix memory_coloring pass when MIGRAPHX_NSTREAMS > 2 #3757

Draft
wants to merge 1 commit into
base: develop
Choose a base branch
from

Conversation

kahmed10
Copy link
Collaborator

I noticed allocation segments reaching close to uint64 max value, which clearly would throw an out of memory error when trying to allocate on the GPU. This happened when MIGRAPHX_NSTREAMS was set to 2 or greater and the model somehow was large enough to trigger it.

Changing from auto to size_t seems to fix the issue.

I noticed allocation segments reaching close to uint64 max value, which clearly would throw an out of memory error when trying to allocate on the GPU. This happened when MIGRAPHX_NSTREAMS was set to 2 or greater and the model somehow was large enough to trigger it.

Changing from `auto` to `size_t` seems to fix the issue.
Copy link

codecov bot commented Jan 11, 2025

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 92.16%. Comparing base (6d02806) to head (2e23878).
Report is 2 commits behind head on develop.

Additional details and impacted files
@@           Coverage Diff            @@
##           develop    #3757   +/-   ##
========================================
  Coverage    92.16%   92.16%           
========================================
  Files          515      515           
  Lines        21978    21978           
========================================
  Hits         20256    20256           
  Misses        1722     1722           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

@TedThemistokleous
Copy link
Collaborator

I'd say pull this out of draft, I've confirmed this works across all the models we were testing this with.

@TedThemistokleous TedThemistokleous self-requested a review January 13, 2025 22:39
@TedThemistokleous TedThemistokleous added the bugfix Fixes a bug found in the code. label Jan 13, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bugfix Fixes a bug found in the code.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants