Move GLM4 f32 attention fix to the correct function#13750
Merged
Conversation
ggerganov
approved these changes
May 24, 2025
Member
|
Huh, that's indeed strange why we didn't get a merge conflict. |
Collaborator
|
Thanks for spotting it quickly |
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
@ggerganov You merged SWA support (#13194) 3 hours before I merged my GLM4 fix (#13639). They touched the same build_attn functions, so there should have been a merge conflict. For whatever reason, my patch was applied to the newly-created
build_attnfunction with a unified_iswa kv cache, which is not used by GLM4. So it didn't work anymore. Here's the fix, moving my patch back to the unified build_attn function... I don't think I've seen something like this before, quite the coincidence.