Adjust sharding rules for ragged_dot operations. #23956
Open
+19
−17
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
![copybara-service[bot]](https://avatars.githubusercontent.com/in/44061?s=60&v=4){kind=small}
copybara-service
bot
force-pushed
the
test_738587239
branch
2 times, most recently
from
b4ead77 to
1ccbf1a
Compare
# Mode 3 (the ragged dimension is a batch dimension) Do not propagate shardings to/from the group_sizes (the 3rd operand) since it is redundant. We will discard this tensor and convert ragged_dot into a standard dot in the partitioner. # Mode 1 (the ragged dimension is a lhs non-contracting dimension) The ragged dimension and the group_size dimension needs full replication in the partitioner. # Mode 2 (the ragged dimension is a contracting dimension) The ragged dimension and the group_size dimension needs full replication in the partitioner. We do not mark the ragged dimension as replicated although it is a contracting dimension. PiperOrigin-RevId: 738587239
copybara-service
bot
force-pushed
the
test_738587239
branch
from
1ccbf1a to
b8c1799
Compare
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Outdated suggestions cannot be applied.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.
Adjust sharding rules for ragged_dot operations.
Mode 3 (the ragged dimension is a batch dimension)
Do not propagate shardings to/from the group_sizes (the 3rd operand) since it is redundant. We will discard this tensor and convert ragged_dot into a standard dot in the partitioner.
Mode 1 (the ragged dimension is a lhs non-contracting dimension)
The ragged dimension and the group_size dimension needs full replication in the partitioner.
Mode 2 (the ragged dimension is a contracting dimension)
The ragged dimension and the group_size dimension needs full replication in the partitioner. We do not mark the ragged dimension as replicated although it is a contracting dimension.