This repository provides a curated and ranked list of Python libraries for machine learning, updated weekly. It aims to help users discover and evaluate high-quality open-source ML tools across various categories, from core frameworks to specialized domains like NLP, computer vision, and time series analysis. The ranking is based on a project-quality score derived from GitHub metrics.

How It Works

The project automatically collects data from GitHub and package managers to calculate a "project-quality score" for each listed library. This score is influenced by metrics such as star count, contributor count, fork count, issue count, package downloads, and last update timestamps. Projects are categorized and ranked, with additional indicators for new, inactive, or trending projects.

Quick Start & Requirements

• Installation: The project itself is a curated list and does not require installation. Users interact with the data directly via the README.
• Dependencies: No specific dependencies are required to view the list. The listed libraries have their own installation requirements, typically pip install <library_name> or conda install -c conda-forge <library_name>.

Highlighted Details

• Comprehensive coverage of 34 ML categories with 920 listed projects.
• Weekly updates ensure the list remains current with trending libraries.
• Detailed metrics (stars, downloads, activity) provided for each project.
• Clear indicators for project health (active, inactive, dead) and recent changes.

Maintenance & Community

The project is actively maintained by the ml-tooling organization. Contributions are welcomed via GitHub issues or pull requests to update projects.yaml.

Licensing & Compatibility

The project itself is licensed under the MIT License. The licenses of the listed libraries vary, with common ones including Apache-2.0, BSD-3-Clause, and MIT. Users should verify individual library licenses for compatibility with their projects.

Limitations & Caveats

The ranking is based on automated metrics, which may not fully capture the nuances of project quality or suitability for specific use cases. Some projects are marked as "dead" (no activity for 12+ months), indicating potential maintenance issues or deprecation.