@flowrag/storage-lancedb
v1.1.3
Published
🔍 Vector storage implementation based on LanceDB. Manages vector embeddings and provides fast semantic search via similarity search. LanceDB is embedded (no server required), supports millions of vectors, and can also work on S3.
Maintainers
zweerReadme
@flowrag/storage-lancedb
Vector storage implementation using LanceDB. Embedded, no server required.
Installation
npm install @flowrag/storage-lancedbUsage
import { LanceDBVectorStorage } from '@flowrag/storage-lancedb';
const vector = new LanceDBVectorStorage({
path: './data/vectors',
dimensions: 384,
});
await vector.upsert([{ id: 'chunk:1', vector: [0.1, 0.2, ...], metadata: {} }]);
const results = await vector.search([0.1, 0.2, ...], 10);
const count = await vector.count();License
MIT