Draft
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This PR introduces a new random vector generation that behaves better than random values and produces high-quality HNSW for generated vectors.
The core idea of better random vectors is based on the generation in a low-dimensional Poincaré Disk. It simulates emdeggings with an internal hierarchy structure.
In addition, it adds a Gaussian noise because in the real world, embeddings have other kinds of internal structure.
As a result, accuracy on params
bfb --dim 1024 --num-vectors 100k --structured-vectorschanges HNSW accuracy test from WebUI from 25% up to 98%. Indexing speed is dramatically decreased because of the quality of the graph (in local test from 35s down to 25s).