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 mpi4py/failing deepspeed test issues #2353

Merged
merged 7 commits into from
Jan 18, 2024
Merged

Fix mpi4py/failing deepspeed test issues #2353

merged 7 commits into from
Jan 18, 2024

Conversation

muellerzr
Copy link
Collaborator

What does this PR do?

Solves the issue of #2350, mpi4py needed to be conda installed in the image, so this PR updates the docker image to do so (since currently we don't have docker images sanctioned by subject and it's just one with everything)

You can see it passing here: https://github.com/huggingface/accelerate/actions/runs/7571310090

Fixes # (issue)

Failing nightly test

Before submitting

  • This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
  • Did you read the contributor guideline,
    Pull Request section?
  • Was this discussed/approved via a Github issue or the forum? Please add a link
    to it if that's the case.
  • Did you make sure to update the documentation with your changes? Here are the
    documentation guidelines, and
    here are tips on formatting docstrings.
  • Did you write any new necessary tests?

Who can review?

Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.

@pacman100 @SunMarc

@HuggingFaceDocBuilderDev

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

Copy link
Member

@SunMarc SunMarc left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for fixing this ! Why do we need conda install ? It is possible to do apip install? I don't like mixing pip and conda. However, I think it is fine as long as you use pip after conda. See related doc.

@muellerzr
Copy link
Collaborator Author

@SunMarc it's a very specific mpi4py thing because the conda install will get some needed compilation dependencies while pip does not. Sourab and I had to find this out the hard way a number of times 😓

@muellerzr
Copy link
Collaborator Author

We also do conda install pip so that way pip envs are stuck in the conda env it was made in

@muellerzr muellerzr changed the title Fix mpy4pi/failing deepspeed test issues Fix mpi4pi/failing deepspeed test issues Jan 18, 2024
@muellerzr muellerzr changed the title Fix mpi4pi/failing deepspeed test issues Fix mpi4py/failing deepspeed test issues Jan 18, 2024
Copy link
Contributor

@pacman100 pacman100 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you @muellerzr for fixing this difficult to debug issue, great investigation. ✨

@muellerzr muellerzr merged commit c7d11d7 into main Jan 18, 2024
4 checks passed
@muellerzr muellerzr deleted the mpi4py branch January 18, 2024 18:01
statelesshz pushed a commit to statelesshz/accelerate that referenced this pull request Jan 22, 2024
* Try deepspeed after installing mpi4py

* Try again

* Just GPU needed

* Run slow deepspeed

* Fin

* Uncomment

* Uncomment x2
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.

None yet

4 participants