This repository has been archived by the owner on Jul 1, 2024. It is now read-only.
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Outdated suggestions cannot be applied.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:
This adds support for stateful losses with learned parameters. Examples include the NormFace / SphereFace / CosFace / ArcFace family of losses, or noise-contrastive estimation losses that learn partition parameters.
Loss parameters are added to the optimizer, and saved to checkpoints. Losses with parameters are wrapped in a separate DistributedDataParallel instance to synchronize parameter initialization, and gradients, across nodes.
Differential Revision: D19717056