Move torchmetrics to device when using FSDP #18954
Merged
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
awaelchli
added
bug
for more information, see https://pre-commit.ci
awaelchli
commented
Nov 6, 2023
…ng into bugfix/metrics-fsdp
for more information, see https://pre-commit.ci
…ng into bugfix/metrics-fsdp
awaelchli
requested review from
carmocca,
justusschock,
Borda and
williamFalcon
as code owners
Codecov Report
Additional details and impacted files@@ Coverage Diff @@
## master #18954 +/- ##
=========================================
- Coverage 75% 51% -24%
=========================================
Files 450 445 -5
Lines 36150 36145 -5
=========================================
- Hits 27247 18476 -8771
- Misses 8903 17669 +8766 |
carmocca
reviewed
Nov 6, 2023
|
I would prefer to first get a comment on the approach #18954 (comment) before doing it. |
carmocca
approved these changes
Nov 8, 2023
Borda
approved these changes
Nov 8, 2023
Borda
pushed a commit
that referenced
this pull request
Nov 14, 2023
(cherry picked from commit 964364b)
lantiga
pushed a commit
that referenced
this pull request
Nov 15, 2023
(cherry picked from commit 964364b)
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Outdated suggestions cannot be applied.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?
Fixes #18888
FSDP doesn't move modules that don't have parameters to the device. TorchMetrics don't have parameters, but require to be move to the right device to handle their state updates on device. This PR checks whether the LightningModule has Metric modules and moves them automatically. This ensures a seamless integration, since metrics are a core feature in Lightning.
Corresponding issue on PyTorch: pytorch/pytorch#113113
📚 Documentation preview 📚: https://pytorch-lightning--18954.org.readthedocs.build/en/18954/
cc @Borda @carmocca @justusschock @awaelchli