musa: enable building fat binaries, enable VMM support, and disable Flash Attention on QY1 (MTT S80) #9526
+29
−5
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Outdated suggestions cannot be applied.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.
This PR enables building fat binaries for both QY1 (mp_21, MTT S80) and QY2 (mp_22, MTT S4000), and restores VMM support (the issue mentioned in #8383 has been fixed). However, due to a known issue when compiling Flash Attention on QY1, it has been explicitly disabled.
Testing done
make GGML_MUSA=1
-> passedran
tinyllama
model on MTT S80 and MTT S4000 -> passed./tests/test-backend-ops
-> passed full log