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

make plugin type check more flexible #20186

Merged
merged 3 commits into from
Nov 26, 2024
Merged

Conversation

jedyang97
Copy link
Contributor

@jedyang97 jedyang97 commented Aug 10, 2024

What does this PR do?

The current type check of plugins for Trainer is restricted to list by isinstance(plugins, list). However, if plugins is instantiated from something like Hydra or OmegaConf, it will have type <class 'omegaconf.listconfig.ListConfig'>, and the current logic would wrap that list-like object with a second-layer list and cause subsequant logic to fail. If you look at the code, it is not necessary that plugins must be a list - it suffices as long as it is iterable. Therefore, the proposed simple fix change the check from isinstance(plugins, list) to isinstance(plugins, Iterable). I hope this can benefit those who uses Hydra or OmegaConf.

Before submitting
  • Was this discussed/agreed via a GitHub issue? (not for typos and docs)
  • Did you read the contributor guideline, Pull Request section?
  • Did you make sure your PR does only one thing, instead of bundling different changes together?
  • Did you make sure to update the documentation with your changes? (if necessary)
  • Did you write any new necessary tests? (not for typos and docs)
  • Did you verify new and existing tests pass locally with your changes?
  • Did you list all the breaking changes introduced by this pull request?
  • Did you update the CHANGELOG? (not for typos, docs, test updates, or minor internal changes/refactors)

PR review

Anyone in the community is welcome to review the PR.
Before you start reviewing, make sure you have read the review guidelines. In short, see the following bullet-list:

Reviewer checklist
  • Is this pull request ready for review? (if not, please submit in draft mode)
  • Check that all items from Before submitting are resolved
  • Make sure the title is self-explanatory and the description concisely explains the PR
  • Add labels and milestones (and optionally projects) to the PR so it can be classified

📚 Documentation preview 📚: https://pytorch-lightning--20186.org.readthedocs.build/en/20186/

@github-actions github-actions bot added the pl Generic label for PyTorch Lightning package label Aug 10, 2024
@jedyang97
Copy link
Contributor Author

Ah, I just realized that one can pass in _convert_: partial in OmegaConf to avoid casting the type to <class 'omegaconf.listconfig.ListConfig'>. In this case, I will leave it to the authors to decide whether this change is still worth it.

Copy link
Collaborator

@lantiga lantiga left a comment

Choose a reason for hiding this comment

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

Approved, completing the PR in #20452

@lantiga lantiga merged commit 13b74f7 into Lightning-AI:master Nov 26, 2024
73 of 74 checks passed
Copy link

codecov bot commented Nov 26, 2024

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 79%. Comparing base (87565cb) to head (a771423).
Report is 2 commits behind head on master.

❗ There is a different number of reports uploaded between BASE (87565cb) and HEAD (a771423). Click for more details.

HEAD has 94 uploads less than BASE
Flag BASE (87565cb) HEAD (a771423)
cpu 46 24
lightning 35 18
pytest 24 0
python3.9 12 6
lightning_fabric 7 0
python3.11 12 6
python3.10 6 3
gpu 2 0
python3.12 16 9
Additional details and impacted files
@@            Coverage Diff            @@
##           master   #20186     +/-   ##
=========================================
- Coverage      88%      79%     -9%     
=========================================
  Files         267      264      -3     
  Lines       23274    23220     -54     
=========================================
- Hits        20381    18271   -2110     
- Misses       2893     4949   +2056     

tshu-w pushed a commit to tshu-w/pytorch-lightning that referenced this pull request Dec 4, 2024
lantiga added a commit that referenced this pull request Dec 10, 2024
* Add `convert_module` to FSDP

* Update ChangeLog

* make plugin type check more flexible (#20186)

Co-authored-by: Jirka Borovec <[email protected]>
Co-authored-by: Luca Antiga <[email protected]>

* Make plugin type check more flexible (Fabric) (#20452)

* make plugin type check more flexible

* Change signature and make the equivalent changes to Fabric connector

---------

Co-authored-by: Jianing Yang <[email protected]>
Co-authored-by: Jirka Borovec <[email protected]>

* Pin setuptools for gpu builds

* Fix link in doc

---------

Co-authored-by: Luca Antiga <[email protected]>
Co-authored-by: Jianing Yang <[email protected]>
Co-authored-by: Jirka Borovec <[email protected]>
Co-authored-by: Luca Antiga <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
pl Generic label for PyTorch Lightning package
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants