[WIP] Basic system check for troubleshooting multi-GPU issues #19609
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
for more information, see https://pre-commit.ci
for more information, see https://pre-commit.ci
…ing into feature/system-check
for more information, see https://pre-commit.ci
awaelchli
added
the
fun
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What does this PR do?
WIP
Users frequently contact us about their multi-GPU training not working (showing as "hanging" or "stuck"). In most cases, this is because of a bad multi-GPU setup. This PR adds a PyTorch-only test script that troubleshoots issues and runs tests. Right now, it only runs a simple all-reduce test and whether P2P is an issue. Future iterations can do more checks. Run it like so:
For future bug reports or channel questions, we can ask users to run this check. For example:
https://pytorch-lightning.slack.com/archives/CRBLFHY79/p1712878650285279
Issues this might close:
#17389
#18490
#19606
#19487
📚 Documentation preview 📚: https://pytorch-lightning--19609.org.readthedocs.build/en/19609/