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.

Sign up for GitHub

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

Jump to bottom

Bump pytorch to 2.5.0 #3657

Closed
wants to merge 1 commit into from
Closed

Bump pytorch to 2.5.0 #3657

wants to merge 1 commit into from

Conversation

b-chu
Copy link
Contributor

@b-chu b-chu commented Oct 17, 2024

What does this PR do?

Bumps pytorch version to 2.5.0

What issue(s) does this change relate to?

Before submitting

  • Have you read the contributor guidelines?
  • Is this change a documentation change or typo fix? If so, skip the rest of this checklist.
  • Was this change discussed/approved in a GitHub issue first? It is much more likely to be merged if so.
  • Did you update any related docs and document your change?
  • Did you update any related tests and add any new tests related to your change? (see testing)
  • Did you run the tests locally to make sure they pass?
  • Did you run pre-commit on your change? (see the pre-commit section of prerequisites)

@b-chu b-chu requested a review from a team as a code owner October 17, 2024 20:26
@b-chu b-chu force-pushed the pytorch2.5 branch 6 times, most recently from d350a83 to da8d2f8 Compare October 18, 2024 14:47
@b-chu b-chu requested a review from a team as a code owner October 18, 2024 14:47
@b-chu b-chu enabled auto-merge (squash) October 18, 2024 15:15
mvpatel2000
mvpatel2000 reviewed Oct 18, 2024
View reviewed changes
Copy link
Contributor

@mvpatel2000 mvpatel2000 left a comment

Choose a reason for hiding this comment

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

you will need to update patch_pytorch as well, e.g for the sync hook for auto microbatching

.github/workflows/pr-cpu.yaml Show resolved Hide resolved
composer/checkpoint/state_dict.py Show resolved Hide resolved
composer/models/huggingface.py Show resolved Hide resolved
docker/build_matrix.yaml Show resolved Hide resolved
tests/checkpoint/helpers.py Show resolved Hide resolved
@b-chu b-chu changed the title Bump pytorch to 2.5.0 [NOT READY FOR REVIEW] Bump pytorch to 2.5.0 Oct 18, 2024
@b-chu b-chu changed the title [NOT READY FOR REVIEW] Bump pytorch to 2.5.0 Bump pytorch to 2.5.0 Oct 18, 2024
@b-chu b-chu requested a review from mvpatel2000 October 18, 2024 21:25
mvpatel2000
mvpatel2000 reviewed Oct 18, 2024
View reviewed changes
Copy link
Contributor

@mvpatel2000 mvpatel2000 left a comment

Choose a reason for hiding this comment

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

what about the SGD and ckpt changes?

docker/generate_build_matrix.py
@@ -180,7 +180,7 @@ def _write_table(table_tag: str, table_contents: str):


def _main():
python_pytorch_versions = [('3.11', '2.4.1'), ('3.11', '2.3.1'), ('3.11', '2.2.2')]
python_pytorch_versions = [('3.11', '2.5.0'), ('3.11', '2.4.1'), ('3.11', '2.3.1')]
Copy link
Contributor

Choose a reason for hiding this comment

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

Suggested change
python_pytorch_versions = [('3.11', '2.5.0'), ('3.11', '2.4.1'), ('3.11', '2.3.1')]
python_pytorch_versions = [('3.12', '2.5.0'), ('3.11', '2.4.1'), ('3.11', '2.3.1')]

3.12?

Copy link
Contributor Author

Choose a reason for hiding this comment

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

Approve 2.5.0 first then I'll make 3.12 changes (same PR) and see how many things it breaks

composer/trainer/_patch_pytorch.py Outdated Show resolved Hide resolved
mvpatel2000
mvpatel2000 reviewed Oct 18, 2024
View reviewed changes
composer/trainer/trainer.py Outdated Show resolved Hide resolved
composer/trainer/trainer.py Show resolved Hide resolved
@b-chu b-chu requested a review from mvpatel2000 October 18, 2024 21:53
@bigning
Copy link
Contributor

bigning commented Oct 18, 2024 •
edited
Loading

the checkpointing change looks good to me. can you run a daily test ?

mvpatel2000
mvpatel2000 reviewed Oct 19, 2024
View reviewed changes
Copy link
Contributor

@mvpatel2000 mvpatel2000 left a comment

Choose a reason for hiding this comment

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

LGTM. Let's do python bump

@b-chu b-chu closed this Oct 19, 2024
auto-merge was automatically disabled October 19, 2024 01:08

Pull request was closed

@b-chu b-chu deleted the pytorch2.5 branch October 19, 2024 01:09
@b-chu
Copy link
Contributor Author

b-chu commented Oct 19, 2024

All daily tests pass
https://github.com/mosaicml/composer/actions/runs/11412854427

@b-chu
Copy link
Contributor Author

b-chu commented Oct 19, 2024

Moving PR to #3663

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@mvpatel2000 mvpatel2000 mvpatel2000 left review comments

@bigning bigning Awaiting requested review from bigning

@dakinggg dakinggg Awaiting requested review from dakinggg

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@b-chu @bigning @mvpatel2000