[BugFix] Fix flash_attn_with_kvcache with scalar cache_seqlen #1795
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
stepinto
force-pushed
the
cache_seqlens_0805
branch
from
9c1843f to
0f5288c
Compare
stepinto
changed the title
When the parameter `cache_seqlen` is scalar, it should be expand to vector of shape (batch_size). In the original code, whenever `block_table` is used, the shape of `k_cache` is (num_blocks, page_size, ...), and thus `cache_seqlen` is expanded to shape (num_blocks) instead of (batch_size), which is wrong. This fix uses the shape of `q`, which is always `batch_size`.
stepinto
force-pushed
the
cache_seqlens_0805
branch
from
0f5288c to
9057ef4
Compare
Member
|
Thank you! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
When the parameter
cache_seqlenis scalar, it should be expand to vector of shape (batch_size). In the original code, wheneverblock_tableis used, the shape ofk_cacheis (num_blocks, page_size, ...), and thuscache_seqlenis expanded to shape (num_blocks) instead of (batch_size), which is wrong. This fix uses the shape ofq, which is alwaysbatch_size.