Initial Release
HNSW4Hashes is a specialized implementation of the HNSW (Hierarchical Navigable Small World) data structured adapted for efficient nearest neighbor search of approximate matching hashes.
Features
- Construction of the HNSW data structure.
- Insertion of nodes into the data strcture.
- K-nearest neighbor search based on similarity.
- Percentage search to retrieve nodes that meet a similarity threshold.
- Several scripts to benchmark the data structure and generate plots.
- Logging functionality for debugging and monitoring.
- Ability to save and restore the index structure