Fix indexing bug when using parallelism to build CPU hierarchy in HNSW #620
Conversation
This file contains 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
|
Build times before bug was fixed: Built times with this PR: |
cjnolet
approved these changes
Jan 27, 2025
divyegala
added
bug
|
@divyegala does this similar problem exist with Cagra index getting converted to Faiss enabled HNSW? |
|
@navneet1v no, this bug is not present in FAISS |
cjnolet
approved these changes
Jan 28, 2025
cjnolet
approved these changes
Jan 30, 2025
|
/merge |
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Outdated suggestions cannot be applied.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.
hnswlib uses an internal indexing system which assigns an ID to points, atomically, in-order that they are added to the index. When using parallelism to add points to the index, the internal ID may be different than the "label" of the point (label, for us, is just the index of the row in the dataset) as a consequence of adding points in-parallel in no deterministic order.
The bug was that I was using the label itself to write out the CPU hierarchy, when I should have been using hnswlib's internal ID for the point associated with that label.