NVIDIA / NVIDIA/cuvs

[FEA] Allow NN Descent to keep input data type for distance computation

Open
#2,085 0 comments 0 reactions 1 assignee View on GitHub

@jinsolp is already working on this.

Since May 14, 2026.

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

Description

Default NN Descent to computing distance in its native type. Need to support local_join_kernel with uint8 and int8 computation.
Add an option for users to configure distance computation dtype that overrides the default.

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.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.