A few weeks back, we released pg_embedding[1], a new extension for Postgres and LangChain which introduced Hierarchical Navigable Small Worlds (HNSW) indexes for vector similarity search. This new indexing method resulted in 20x faster queries at a 99% accuracy compared to traditional IVFFlat indexing[2].
Today, we released a new version which includes the following improvements:
1. The HNSW index is now constructed on disk instead of in memory
2. The extension now supports Cosine, Manhattan, and Euclidean distances
1 comment
[ 2.1 ms ] story [ 9.4 ms ] threadToday, we released a new version which includes the following improvements:
1. The HNSW index is now constructed on disk instead of in memory
2. The extension now supports Cosine, Manhattan, and Euclidean distances
[1]: https://github.com/neondatabase/pg_embedding
[2]: https://neon.tech/blog/pg-embedding-extension-for-vector-sea...