[TOSA] MultiheadAttention legalization #4382
Open
+304
−20
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
catcor01
force-pushed
the
multihead_attention
branch
2 times, most recently
from
a98526f to
cf45a2e
Compare
Collaborator
Lallapallooza
left a comment
Lallapallooza
left a comment
There was a problem hiding this comment.
Thanks for patch, few comments.
projects/pt1/test/python/scaled_dot_product_attention_lowering.py
Outdated
Show resolved
Hide resolved
- Legalize Torch scaled_dot_product_attention into TOSA by adding the necessary patterns in TorchToTosa.cpp plus backend type-conversion hooks. - Introduce a detailed decomposition path for multi-head attention within DecomposeComplexOps.cpp, preparing inputs for TOSA lowering. - Expands the PT1 e2e suite with a dedicated multi-head attention MLIR/Python test and drop the corresponding xfails now that the path works. Signed-off-by: Cathal Corbett <cathal.corbett@arm.com> Change-Id: I96c17aefd25b979f1cf6e897d91d5a29f0a2fa85
catcor01
force-pushed
the
multihead_attention
branch
from
cf45a2e to
fd02d37
Compare
sahas3
reviewed
Dec 4, 2025
| PatternRewriter &rewriter); | ||
|
|
||
| namespace { | ||
| // Decompose scaled dot product attention into matmul/softmax pipeline when |
Member
There was a problem hiding this comment.
Is this decomposition producing any different IR compared to leveraging the decomposition of sdpa with ExportedProgram.run_decompositions https://docs.pytorch.org/docs/stable/export.html#export-ir-decompositions -- see https://discord.com/channels/636084430946959380/742573221882364009/1446121930922004623 for reference.
I am wondering if the sdpa op should be added to the default decomposition list in
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.
No description provided.