fix: for balanced kmeans use grid.x for adjust_centers to avoid grid.y overflow #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
…y overflow for large n_clusters
|
@Nischal1729 It seems the original repository would benefit from this fix - I would recommend raising the same PR to |
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Outdated suggestions cannot be applied.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.
fix: for balanced kmeans use grid.x for adjust_centers to avoid grid.y overflow for large n_clusters
Problem: adjust_centers launched with grid.y = ceil(n_clusters/4) hits CUDA’s 65,535 Y-dim limit for n_clusters > 262,140 (e.g., 263k and 1M).
Fix: enumerate blocks along grid.x and compute l via blockIdx.x. No algorithmic or perf changes; only prevents the invalid configuration.
Repro: 262k works; 263k fails pre-fix; both work post-fix. 1M centroid training proceeds past balancing.
Impact: zero regression; removes a hard cap on n_clusters.