Fast and memory-efficient approximate nearest neighbor search with JAX
Project description
Annax: Approximate Nearest Neighbor Search with JAX
Annax is a high-performance Approximate Nearest Neighbor (ANN) library built on top of the JAX framework. It provides fast and memory-efficient search for high-dimensional data in various applications, such as large-scale machine learning, computer vision, and natural language processing. Annax leverages the power of GPU acceleration to deliver outstanding performance and includes a wide range of indexing structures and distance metrics to cater to different use cases. The easy-to-use API makes it accessible to both beginners and experts in the field.
Features
- Fast and memory-efficient approximate nearest neighbor search
- GPU acceleration for high-performance computing
- Supports a wide range of indexing structures and distance metrics
- Easy-to-use API for seamless integration with existing projects
- Applicable to various domains, including machine learning, computer vision, and natural language processing
- Built on top of the JAX framework for enhanced flexibility and extensibility
Installation
To install Annax, simply run the following command in your terminal:
pip install annax
Quick Start
Here's a simple example of using Annax to find the nearest neighbors in a dataset:
import numpy as np
import annax
# Generate some random high-dimensional data
data = np.random.random((1000, 128))
# Create an Annax index with the default configuration
index = annax.Index(data)
# Query for the 10 nearest neighbors of a random vector
query = np.random.random(128)
neighbors, distances = index.search(query, k=10)
Index Types
annax.Index
: Flat Indexannax.IndexIVF
: Inverted File Indexannax.IndexPQ
: Product Quantization Indexannax.IndexIVFPQ
: Inverted File Index with Product Quantization
Development
To install Annax for development, run the following commands in your terminal:
python -m pip install -e '.[dev]'
pre-commit install
License
Annax is released under the MIT License.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file annax-0.0.1.tar.gz
.
File metadata
- Download URL: annax-0.0.1.tar.gz
- Upload date:
- Size: 13.5 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.1 CPython/3.11.3
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 8fc2bfc832404d8a596daa485e5bdce3003e9ce9f756e795533ebcaa97c16670 |
|
MD5 | 44681d8d22f2a3190fc32849bb231452 |
|
BLAKE2b-256 | 5c40386479c008750000e79f8d828c83fd336032522cb3d22c42a25a8018fe5c |
File details
Details for the file annax-0.0.1-py3-none-any.whl
.
File metadata
- Download URL: annax-0.0.1-py3-none-any.whl
- Upload date:
- Size: 12.2 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.1 CPython/3.11.3
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | a0ee37539dab61b739b909b39817251f7eca1f636b8d8199c2c3b37e1b469e4b |
|
MD5 | f52c38e23a82e72e37553360b4928cbf |
|
BLAKE2b-256 | 73d233108305487360cfd2ce0da2705ca1648ae64d08fbf368fc109dc627a804 |