Reduce repeat_interleave calls in apply_F_to_columns#69
Merged
Conversation
Compute cols_rep once and derive idx_in_col / flat_indices via prefix[cols_rep] and starts[cols_rep] indexing. Drops two repeat_interleave calls and a torch.cat per bucket. Add direct unit-test coverage for apply_F_to_columns. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Aida-Rahmattalabi
approved these changes
May 13, 2026
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
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Refactors the column-indexing math in
apply_F_to_columnsso thatcols_rep = arange(K).repeat_interleave(lengths)is computed once and reused to deriveprefix_repandoffsvia indexing. Drops 2repeat_interleavecalls and atorch.catper bucket.Also adds a
TestApplyFToColumnstest class covering identity, scaling, and multi-bucket equivalence, since this path had no direct unit tests before.Testing
7.5% reduction in peak memory