vecmindb 0.1.1

High-performance vector database with auto-tuning capabilities

VecminDB

A high-performance vector database with multiple index algorithms, optimizers, and auto-tuning capabilities.

Features

• Multiple Index Algorithms: HNSW, IVF, PQ, LSH, VPTree, ANNOY, NGT, and more
• Multi-Objective Optimization: NSGA-II, MOEAD, MOPSO for index tuning
• Auto-Tuning: Automatic parameter optimization for best performance
• Parallel Processing: Built-in support for parallel vector operations
• Flexible Storage: Sled-based persistent storage with transaction support
• Caching System: Multi-tier caching with Redis support
• Resource Management: Intelligent memory and CPU resource allocation
• Monitoring: Built-in performance metrics and query statistics
• Multi-Language Support: First-class support for Rust and Python, plus HTTP API

Quick Start

As a Library

use vecmindb::{VectorDB, IndexType, SimilarityMetric};

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    // Create a new vector database
    let db = VectorDB::new("./data")?;
    
    // Create a collection with 128-dimensional vectors using HNSW index
    let collection = db.create_collection("my_vectors", 128, IndexType::HNSW)?;
    
    // Add vectors (collection is Arc<RwLock<VectorCollection>>)
    // Note: RwLock is from tokio::sync, so we use .await for lock acquisition
    {
        let mut coll = collection.write().await;
        coll.add_vector("vec1", &vec![0.1; 128], None).await?;
        coll.add_vector("vec2", &vec![0.2; 128], None).await?;
    }
    
    // Search for similar vectors
    let results = {
        let coll = collection.read().await;
        coll.search(
            &vec![0.15; 128], 
            10, 
            SimilarityMetric::Cosine
        ).await?
    };
    
    for result in results {
        println!("ID: {}, Score: {}", result.id, result.score);
    }
    
    Ok(())
}

As an HTTP Server

# Start the server
cargo run --bin vecmindb-server --features http-server

# Create a collection
curl -X POST http://localhost:8080/api/v1/collections \
  -H "Content-Type: application/json" \
  -d '{"name": "my_vectors", "dimension": 128, "index_type": "HNSW"}'

# Add a vector
curl -X POST http://localhost:8080/api/v1/collections/my_vectors/vectors \
  -H "Content-Type: application/json" \
  -d '{"id": "vec1", "values": [0.1, 0.2, ...], "metadata": {}}'

# Search vectors
curl -X POST http://localhost:8080/api/v1/collections/my_vectors/search \
  -H "Content-Type: application/json" \
  -d '{"query": [0.15, 0.25, ...], "top_k": 10, "metric": "cosine"}'

🐳 Docker Deployment

Run VecminDB in seconds with Docker:

# Using Docker Compose (Recommended)
docker-compose up -d

# Using Docker CLI
docker build -t vecmindb .
docker run -p 8080:8080 -v ./data:/data vecmindb

Installation

Add this to your Cargo.toml:

[dependencies]
vecmindb = "0.1"

# Optional features
vecmindb = { version = "0.1", features = ["http-server", "distributed"] }

Python Bindings 🐍

To use VecminDB in Python, please install from source (until v0.1.0 is on PyPI):

pip install maturin
cd bindings/python
maturin develop --release

See INSTALL_PYTHON.md for details.

Supported Index Types

• HNSW: Hierarchical Navigable Small World graphs
• IVF: Inverted File Index
• PQ: Product Quantization
• LSH: Locality-Sensitive Hashing
• VPTree: Vantage-Point Tree
• ANNOY: Approximate Nearest Neighbors Oh Yeah
• NGT: Neighborhood Graph and Tree
• Flat: Brute-force exact search

Performance

VecminDB is designed for high performance:

• Parallel vector operations using Rayon
• SIMD optimizations (optional)
• Efficient memory management
• Smart caching strategies
• Auto-tuning for optimal parameters

Documentation

For detailed documentation, see:

• API Documentation
• User Guide
• Examples

🐍 Python Bindings

VecminDB is available on PyPI!

pip install vecmindb

👉 Quick Start Guide - Get started in 5 minutes. 📚 Build from Source - Instructions for development.

License

Licensed under either of:

• Apache License, Version 2.0 (LICENSE-APACHE)
• MIT license (LICENSE-MIT)

at your option.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.