A lightweight research library for managing Approximate Nearest Neighbor search datasets.
A vector bulk data processing tool
Rust implementation of vectordata tools for dataset.yaml
A streamable, readable, writeable, randomly accessible file format
Approximate nearest-neighbor search
A Rust implementation of DiskANN (Disk-based Approximate Nearest Neighbor search) using the Vamana graph algorithm. Provides memory-efficient vector search through graph traversal and memory-mapped storage, enabling billion-scale search with minimal RAM usage.
Neuros is a Rust package for Artificial (Feedforward) Neural Networks (ANNs) processing.
Advanced vector search: RaBitQ quantization with IVF and MSTG (Multi-Scale Tree Graph) index
IMECE v0.1.0 - Local-First Autonomous Agent Framework. Zero API dependency. Zero cloud lock-in. Full sovereignty.
Standalone vector data I/O — read and write fvec, ivec, mvec, dvec, bvec, svec, npy, parquet, and slab formats
The universal annotation converter
Embedding methods implemented in Rust: (parametric) UMAP, tSNE, PHATE, Diffusion Map and PacMAP.