fixed matmul broadcasting #25663
Open
fixed matmul broadcasting #25663
+19
−1
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
Signed-off-by: Konstantin Pueckler <[email protected]>
|
@microsoft-github-policy-service agree company="QuantCo" |
|
@xadupre might you be in a position to review this? |
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Outdated suggestions cannot be applied.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.
Description
currently, the broadcasting of the
MatMuloperator does not correctly handle zero-sized batch dimensions. By the numpy standard, when one matrix has length 0 in one batch dimension and the other matrix has length 1 in that dimension, it should be broadcasted to length 0. Instead, onnxruntime currently raises an error in that case.I modified the computation of
output_dimsto fix that error.Motivation and Context
We came across this problem when testing the onnxruntime via onnx-tests.