perf: faster merge counting — single-pass dict + inline max#25
Merged
Conversation
AmitMY
force-pushed
the
perf/faster-counting
branch
2 times, most recently
from
c197e10 to
1b98e73
Compare
- Replace Counter with plain dict for merge counting (~faster hashing) - Find best merge in single pass instead of building a second Counter and calling most_common(1) (avoids heap sort) - Remove redundant ONLY_MINIMAL_MERGES filter from trainer — Tree merges are now filtered at the source in Tree.get_merges() - Drop unused Counter import 100 samples / 200 merges: 3.03s → 2.90s (~5% faster) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
AmitMY
force-pushed
the
perf/faster-counting
branch
from
1b98e73 to
5781747
Compare
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Counterwith plaindictfor merge countingCounter+most_common(1)(heap sort)ONLY_MINIMAL_MERGESfilter from trainer —Tree.get_merges()now filters at the sourceCounterimportPerformance
~5% improvement. The remaining bottleneck is the full graph rescan each iteration — fixing that would require incremental counting (only update positions affected by a merge), which would be a larger architectural change.
Test plan
ruff check .passes🤖 Generated with Claude Code