⚡ Build powerful extensions for the world's most efficient tabular foundation model ⚡
Looking for the main TabPFN package? Check out TabPFN.
# Clone and install the repository
git clone https://github.com/priorlabs/tabpfn-extensions.git
pip install -e tabpfn-extensions
# Choose one of the following installation options:
# 1. For GPU-accelerated local inference:
pip install tabpfn
# 2. For cloud-based inference via API:
pip install tabpfn-client
Choose the right TabPFN implementation for your needs:
- TabPFN Client: Easy-to-use API client for cloud-based inference
- TabPFN Extensions (this repo): Community extensions and integrations
- TabPFN: Core implementation for local deployment and research
TabPFN Extensions is a collection of community-driven extensions and tools built around TabPFN, the state-of-the-art foundation model for tabular data. This repository makes it easy to:
- Build domain-specific extensions
- Create integrations with other frameworks
- Share utilities and tools
- Contribute example applications
- Develop custom solutions
Here are some highlighted community extensions:
🔮 Unsupervised Learning
- Data generation capabilities
- Outlier detection
- Distribution modeling
🔍 Interpretability
- Feature importance analysis
- Model explanation tools
- Decision boundary visualization
⚡ AutoTabPFN
- Post-hoc ensemble techniques
- Automatic hyperparameter tuning
- Optimized performance
🌲 Random Forest PFN
- Random forest adaptation of TabPFN
- Scalable for larger datasets
- Parallel processing support
And many more! Browse the full list of extensions.
Each extension lives in its own subpackage:
tabpfn-extensions/
├── src/
│ └── tabpfn_extensions/
│ └── your_package/ # Your extension code
├── examples/
│ └── your_package/ # Usage examples
├── tests/
│ └── your_package/ # Tests
└── requirements/
└── your_package.txt # Dependencies (optional)
We welcome all contributions! See our Contributing Guide for details.
Quick start:
- Fork the repository
- Create your package under
src/
- Add examples
- Submit a PR
This project is licensed under the Apache License 2.0 - see the LICENSE.txt file for details.
Built with ❤️ by the TabPFN community