Add find_intermediate_sharding to optimize resharding. #146
Conversation
This file contains hidden or 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
This change introduces `find_intermediate_sharding`, which attempts to find an intermediate sharding to optimize resharding between two NamedShardings. This is particularly useful when resharding from a sharding with a larger total number of shards to one with a smaller total number of shards, where a direct resharding might involve an expensive all-gather. The intermediate sharding is constructed by splitting certain sharded dimensions into a "split" axis and a "replica" axis, allowing for a more efficient two-step resharding process. Helper functions are added to analyze sharding dimensions and build the intermediate mesh and PartitionSpec. Tests are included for the new functionality. PiperOrigin-RevId: 846807164
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.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.
Add find_intermediate_sharding to optimize resharding.
This change introduces
find_intermediate_sharding, which attempts to find an intermediate sharding to optimize resharding between two NamedShardings. This is particularly useful when resharding from a sharding with a larger total number of shards to one with a smaller total number of shards, where a direct resharding might involve an expensive all-gather. The intermediate sharding is constructed by splitting certain sharded dimensions into a "split" axis and a "replica" axis, allowing for a more efficient two-step resharding process. Helper functions are added to analyze sharding dimensions and build the intermediate mesh and PartitionSpec. Tests are included for the new functionality.