pytorch / pytorch/vision

unittests on Linux GPU often run out of memory

Open
#7,736 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

module: ci
Dominant language
Python
Stars
17.9k
Forks
7.3k
Avg merge
1d 15h
Merged PRs (30d)
13

Description

For example: https://github.com/pytorch/vision/actions/runs/5485581154/attempts/2#summary-14857209658

Looking at the HUD this failure seems to have started around June 23.

Nothing on our side has changed around that time that could cause this. Plus, there are runs where this job succeeds: https://github.com/pytorch/vision/actions/runs/5400569293/attempts/1#summary-14621027485

cc @seemethere

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Compare the failing and successful GitHub Actions runs linked in the issue, then inspect the HUD history for the unittests-linux job around June 23. Trace which Linux GPU tests or resource conditions lead to out-of-memory failures; done means the job completes reliably without running out of memory.

Written by the indexing model from the issue text.

Assessment

Tech stack
github-actions, linux, python
Domain
ci-cd, infrastructure, testing
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.