-
Notifications
You must be signed in to change notification settings - Fork 3.4k
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
Maintain float32 precision at minimum in ResultMetric #18686
Merged
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
for more information, see https://pre-commit.ci
awaelchli
added
bug
Something isn't working
logging
Related to the `LoggerConnector` and `log()`
labels
Oct 2, 2023
awaelchli
changed the title
Maintain float32 precision or higher in ResultMetric
Maintain float32 precision at minimum in ResultMetric
Oct 2, 2023
…fix/float16-logging
awaelchli
requested review from
carmocca,
justusschock,
Borda and
williamFalcon
as code owners
October 2, 2023 11:28
carmocca
approved these changes
Oct 2, 2023
src/lightning/pytorch/trainer/connectors/logger_connector/result.py
Outdated
Show resolved
Hide resolved
Codecov Report
Additional details and impacted files@@ Coverage Diff @@
## master #18686 +/- ##
==========================================
- Coverage 83% 50% -33%
==========================================
Files 428 420 -8
Lines 33546 33405 -141
==========================================
- Hits 27985 16824 -11161
- Misses 5561 16581 +11020 |
Borda
approved these changes
Oct 2, 2023
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Outdated suggestions cannot be applied.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 #18620
Fixes #17414 ?
This PR ensures that reduction of floating point values happens in float32 precision at least (the user can also choose to operate in higher (double) precision). The PR ensures that if you pass in float16/bfloat16 values, they get upcast to float32 for calculations inside the metric, even if
torch.get_default_dtype()
is lower precision.Before submitting
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
📚 Documentation preview 📚: https://pytorch-lightning--18686.org.readthedocs.build/en/18686/
cc @Borda @carmocca @Blaizzy