Generic and Open Learning Federator

A scalable, portable, and lightweight Federated Learning framework.

• Free software: MIT license
• Documentation: https://generic-and-open-learning-federator.readthedocs.io.

Citations

If this project is helpful to your research, please cite our papers:

L. You, Z. Guo, C. Yuen*, C.Y.C. Chen, Y. Zhang, H.V. Poor,"A framework reforming personalized Internet of Things by federated meta-learning", Nature Communications, 2025.

L. You, Z. Guo, B. Zuo, Y. Chang*, C. Yuen,"SLMFed: A Stage-based and Layer-wise Mechanism for Incremental Federated Learning to Assist Dynamic and Ubiquitous IoT", IEEE Internet of Things Journal, 2024.

S. Liu, L. You*, R. Zhu, B. Liu, R. Liu, Y. Han, C. Yuen,"AFM3D: An Asynchronous Federated Meta-learning Framework for Driver Distraction Detection", IEEE Transactions on Intelligent Transportation Systems, 2024.

L. You, S. Liu, B. Zuo, C. Yuen*, D. Niyato, H. V. Poor,"Federated and Asynchronized Learning for Autonomous and Intelligent Things", IEEE Network Magazine, 2023.

L. You, S. Liu, T. Wang, B. Zuo, Y. Chang, C. Yuen*,"AiFed: An Adaptive and Integrated Mechanism for Asynchronous Federated Data Mining", IEEE Transactions on Knowledge and Data Engineering, 2023.

L. You, S. Liu, Y. Chang, C. Yuen*,"A triple-step asynchronous federated learning mechanism for client activation, interaction optimization, and aggregation enhancement", IEEE Internet of Things Journal, 2022.

@article{You2025framework,
  title={A framework reforming personalized Internet of Things by federated meta-learning},
  author={You, Linlin and Guo, Zihan and Yuen, Chau and Chen, Calvin Yu-Chian and Zhang, Yan and Poor, H. Vincent},
  journal={Nature communications},
  volume={16},
  pages={3739},
  year={2025},
  publisher={Nature Publishing Group UK London}
}

@article{you2024slmfed,
  title={SLMFed: A Stage-Based and Layerwise Mechanism for Incremental Federated Learning to Assist Dynamic and Ubiquitous IoT},
  author={You, Linlin and Guo, Zihan and Zuo, Bingran and Chang, Yi and Yuen, Chau},
  journal={IEEE Internet of Things Journal},
  volume={11},
  number={9},
  pages={16364--16381},
  year={2024},
  publisher={IEEE}
}

@article{liu2024afm3d,
  title={AFM3D: An asynchronous federated meta-learning framework for driver distraction detection},
  author={Liu, Sheng and You, Linlin and Zhu, Rui and Liu, Bing and Liu, Rui and Yu, Han and Yuen, Chau},
  journal={IEEE Transactions on Intelligent Transportation Systems},
  volume={25},
  number={8},
  pages={9659--9674},
  year={2024},
  publisher={IEEE}
}

@article{you2023federated,
  title={Federated and asynchronized learning for autonomous and intelligent things},
  author={You, Linlin and Liu, Sheng and Zuo, Bingran and Yuen, Chau and Niyato, Dusit and Poor, H Vincent},
  journal={IEEE Network},
  volume={38},
  number={2},
  pages={286--293},
  year={2023},
  publisher={IEEE}
}

@article{you2023aifed,
  title={AiFed: An adaptive and integrated mechanism for asynchronous federated data mining},
  author={You, Linlin and Liu, Sheng and Wang, Tao and Zuo, Bingran and Chang, Yi and Yuen, Chau},
  journal={IEEE Transactions on Knowledge and Data Engineering},
  volume={36},
  number={9},
  pages={4411--4427},
  year={2023},
  publisher={IEEE}
}

@article{you2022triple,
  title={A triple-step asynchronous federated learning mechanism for client activation, interaction optimization, and aggregation enhancement},
  author={You, Linlin and Liu, Sheng and Chang, Yi and Yuen, Chau},
  journal={IEEE Internet of Things Journal},
  volume={9},
  number={23},
  pages={24199--24211},
  year={2022},
  publisher={IEEE}
}

Features

• GOLF provides a lightweight solution to support the implementation of FL.
• GOLF modularizes system functions to achieve loose coupling during system development and deployment, which makes the framework more generic and scalable.
• GOLF uses container technology to ensure that the system is weakly dependent on the compilation environment to achieve portability.
• GOLF is compatible with multiple devices (e.g., Android, embedded computers, edge devices, etc.).

News

1. 🌟 June 07, 2024 - Introducing Cedar:

Cedar is a secure, cost-efficient, and domain-adaptive framework for federated meta-learning. Key features include:

• 💡 Federated Meta-Learning: Enable a safeguarded knowledge transfer with high model generalizability and adaptability.
• 📨 Cost-Efficient: Implement a layer-wise model uploading mechanism to reduce communication cost.
• 🔒 Robust Security: Defend against malicious attacks like data inversion and model poisoning.
• 🔧 High Performance: Support high-performance personalization and customization of globally shareable meta-models.

Installation

To install GOLF, simply use pip:

pip install golf_federated

Credits

This package was created with Cookiecutter and the audreyr/cookiecutter-pypackage project template.

Contributing

We welcome contributions! Here are some ways you can help:

1. Report bugs and request features on GitHub Issues: https://github.com/IntelligentSystemsLab/generic_and_open_learning_federator/issues
2. Submit pull requests to improve the codebase.

Contact

For any questions or issues, please contact the development team at guozh29@mail2.sysu.edu.cn.