[Tosa] : Match accumulator type with torch for lowering aten.mm to tosa.matmul #4264
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
sahas3
force-pushed
the
tosaMatmuli8
branch
2 times, most recently
from
a5ec231 to
894b05b
Compare
…ten.add" This reverts commit 0dd01e6.
Lallapallooza
pushed a commit
to Lallapallooza/torch-mlir
that referenced
this pull request
Jul 17, 2025
…sa.matmul (llvm#4264) For `i8` input, accumulator was being set to `i8` as well which produces invalid `tosa.matmul` op. `fp16` input used `fp16` accumulator which while valid by tosa spec, fails numerical verification. This change uses the accumulator type as used by PyTorch (tosa-to-linalg does the same).
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Outdated suggestions cannot be applied.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.
For
i8input, accumulator was being set toi8as well which produces invalidtosa.matmulop.fp16input usedfp16accumulator which while valid by tosa spec, fails numerical verification. This change uses the accumulator type as used by PyTorch (tosa-to-linalg does the same).