Fix matmul flop counting for vector operands - #159
Open
harshitgavita-07 wants to merge 2 commits into
Open
harshitgavita-07 wants to merge 2 commits into
harshitgavita-07 wants to merge 2 commits into
Conversation
meta-cla
Bot
added
the
CLA Signed
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Fixes #130.
matmul_flop_jit()assumed both operands are matrices. With a vector on the right (vector-vector or matrix-vector matmul), the shape assertion indexes[-2]on a 1-D shape and raisesIndexError. With a vector on the left against a batched matrix, the formulaprod(input_shapes[0]) * input_shapes[1][-1]counts only one batch and undercounts.The handler now checks the shared contraction dimension (
-1for a vector right operand,-2otherwise) and computesprod(output_shape) * contraction_dim, which matches the existing behavior for plain matrix-matrix inputs and covers the vector cases.Tests
Added
test_matmul_vectorscovering vector-vector, matrix-vector, batched-matrix-vector, vector-matrix, and vector-batched-matrix. 4 of the 5 cases fail on current main (3 IndexError, 1 wrong count); all pass with the fix. Fulltests/test_flop_count.pysuite passes.