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

Add offloading tests and fix obscure edge case #1860

Merged
merged 5 commits into from
Oct 18, 2024

Conversation

janeyx99
Copy link
Contributor

@janeyx99 janeyx99 commented Oct 17, 2024

Context

What is the purpose of this PR? Is it to

  • add a new feature
  • fix a bug
  • update tests and/or documentation
  • other (please add here)

Please link to any issues this PR addresses.

Changelog

What are the changes made in this PR?

  • Adds basic unit test for offloading for correctness
  • Fixes a really obscure bug that could happen (it has never been reported yet) and adds a test for that

Test plan

Please make sure to do each of the following if applicable to your PR. If you're unsure about any one of these just ask and we will happily help. We also have a contributing page for some guidance on contributing.

  • run pre-commit hooks and linters (make sure you've first installed via pre-commit install)
  • add unit tests for any new functionality
  • update docstrings for any new or updated methods or classes
  • run unit tests via pytest tests
  • run recipe tests via pytest tests -m integration_test
  • manually run any new or modified recipes with sufficient proof of correctness
  • include relevant commands and any other artifacts in this summary (pastes of loss curves, eval results, etc.)

UX

If your function changed a public API, please add a dummy example of what the user experience will look like when calling it.
Here is a docstring example
and a tutorial example

  • I did not change any public API
  • I have added an example to docs or docstrings

Copy link

pytorch-bot bot commented Oct 17, 2024

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/torchtune/1860

Note: Links to docs will display an error until the docs builds have been completed.

✅ No Failures

As of commit c14aa1a with merge base 7d29c21 (image):
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

@facebook-github-bot facebook-github-bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Oct 17, 2024
Copy link
Contributor

@albanD albanD left a comment

Choose a reason for hiding this comment

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

SGTM only small things.

torchtune/training/_activation_offloading.py Outdated Show resolved Hide resolved
torchtune/training/_activation_offloading.py Outdated Show resolved Hide resolved
@codecov-commenter
Copy link

Codecov Report

Attention: Patch coverage is 0% with 7 lines in your changes missing coverage. Please review.

Project coverage is 67.33%. Comparing base (c70ad29) to head (7ad438d).
Report is 11 commits behind head on main.

Files with missing lines Patch % Lines
torchtune/training/_activation_offloading.py 0.00% 7 Missing ⚠️
Additional details and impacted files
@@            Coverage Diff             @@
##             main    #1860      +/-   ##
==========================================
+ Coverage   67.30%   67.33%   +0.03%     
==========================================
  Files         304      305       +1     
  Lines       16000    16034      +34     
==========================================
+ Hits        10768    10797      +29     
- Misses       5232     5237       +5     

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

# memory usage, but this case should not happen often.
unpacked_tensor = self.bwd_tensor_stash[unpack_tensor_id]
if any(
o.untyped_storage() is unpacked_tensor.untyped_storage()
Copy link
Contributor

Choose a reason for hiding this comment

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

When I run with this, I get:

    o.untyped_storage() is unpacked_tensor.untyped_storage()
AttributeError: 'NoneType' object has no attribute 'untyped_storage'

I think o can be None.

Copy link
Contributor Author

Choose a reason for hiding this comment

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

Good catch thanks! Should be updated in latest commit

@janeyx99 janeyx99 changed the title [Edge case bug fix] Use record_stream sometimes for offloading Add offloading tests and fix obscure edge case Oct 18, 2024
Copy link
Contributor

@albanD albanD left a comment

Choose a reason for hiding this comment

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

SGTM!

@janeyx99 janeyx99 marked this pull request as ready for review October 18, 2024 18:53
Copy link
Contributor

@joecummings joecummings left a comment

Choose a reason for hiding this comment

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

applause

@janeyx99 janeyx99 merged commit 3ca0d30 into pytorch:main Oct 18, 2024
17 checks passed
@janeyx99 janeyx99 deleted the offloading-fix branch October 18, 2024 19:00
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

6 participants