NVIDIA / NVIDIA/cuvs

[FEA] Build CAGRA Graph with RaBitQ Dataset View

Open
#2,315 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

feature request
Dominant language
Cuda
Stars
854
Forks
236
Avg merge
3d 3h
Merged PRs (30d)
62

Description

With the new Dataset API, we should now be able to create dataset implementations that store additional state, such as correction codes for Lucene's BBQ quantizers.

This feature will allow users to build a CAGRA graph directly from BBQ (or RaBitQ) quantized vectors.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reading the new Dataset API and the existing CAGRA graph-building path. Determine how a dataset view can retain BBQ or RaBitQ correction codes and how quantized vectors should be passed into graph construction. Done means users can build a CAGRA graph directly from BBQ or RaBitQ quantized vectors.

Written by the indexing model from the issue text.

Assessment

Domain
machine-learning, search
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.