This repository has been archived by the owner on Nov 22, 2022. It is now read-only.
Fix loss reporting #453
Closed
Fix loss reporting #453
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
Summary: When computing average loss for train/eval/test, loss across different batches should be weighted with batch size. This doesn't matter for standard training where most batches are the full batch size except the last. However, when training in a massively distributed fashion (eg: for Federated Learning), data is split into small sizes. Many batches are not 'complete' batches, so weighting by batch size is important. Differential Revision: D14791247 fbshipit-source-id: 5d99ce1adf6e142aba8963e9fca8bbc77787df59
facebook-github-bot
added
the
CLA Signed
|
This pull request has been merged in fb58d20. |
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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Outdated suggestions cannot be applied.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.
Summary:
When computing average loss for train/eval/test, loss across different batches should be weighted with batch size.
This doesn't matter for standard training where most batches are the full batch size except the last. However, when training in a massively distributed fashion (eg: for Federated Learning), data is split into small sizes. Many batches are not 'complete' batches, so weighting by batch size is important.
Differential Revision: D14791247