Get the latest tech news

Show HN: HNSW index for vector embeddings in approx 500 LOC


Heirarchical Navigable Small Worlds. Contribute to dicroce/hnsw development by creating an account on GitHub.

Fast (uses Eigen for SIMD acceleration of distance calculations). When searches arrive they start at the top and search that level (following connections) until they find the closest node in the top level. As the search progresses the code keeps track of the K nearest nodes it has seen.

Get the Android app

Or read this on Hacker News

Read more on:

Photo of LOC

LOC

Photo of vector embeddings

vector embeddings

Photo of approx

approx

Related news:

News photo

Show HN: Python micro event loop library (~250 LOC)

News photo

Lexar doubles up the NM790 series SSDs max capacity to 8TB — new drives spotted at retail for approx $1,000

News photo

Cascading retrieval: Unifying dense and sparse vector embeddings with reranking