Conversation
matmul_hadUt_cuda passes transpose=True to matmul_hadU_cuda, which has no such parameter, so every call raises TypeError. The function has no call sites. Its CPU counterpart matmul_hadUt still works because matmul_hadU does take transpose. Signed-off-by: Anai-Guo <antai12232931@outlook.com>
This branch has not been deployed
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem
matmul_hadUt_cudacannot be called — it forwards atransposekeyword thatmatmul_hadU_cudadoes not accept:Any call raises:
The commented-out
# if transpose:block directly above is the residue:transposesupport was dropped from
matmul_hadU_cuda, but this wrapper was never updated.Why removal rather than restoring
transposematmul_hadUt_cudahas no call sites anywhere in the repo (grepoversrc/,tests/,brevitas_examples/). Onlymatmul_hadU_cudais imported and used —in
graph/equalize.py:1527,nn/equalized_layer.py:110,128, andgraph/hadamard.py:178,181.i.e. resurrecting behaviour that was deliberately disabled. That's a semantic decision
for maintainers, not a bug fix.
caller, in-tree or out.
The CPU counterpart is unaffected and keeps working, since
matmul_hadUdoes taketranspose:Verification
Binding each wrapper's call against its callee's real signature, with the working CPU
pair as a control:
Change
Deletes the four lines of the dead wrapper. No other behaviour touched.
🤖 Generated with Claude Code