Release Notes

MindSpore Federated 0.1 Release Notes

MindSpore Federated is an open source federated learning tool for MindSpor, supports the various participants of machine learning to build AI models together without directly sharing local data. To improve the usability of the federated framework and user development efficiency, related code has been separated from the MindSpore framework and Federated is now an independent repository.

Major Features and Improvements

  • Supports cross-device federation that supports dynamic scaling, disaster recovery, and efficient compression for upload and download traffic.

  • Supports cross-silo federation that can be trained with single-machine multi-card and multi-machine multi-card.

  • Supports vertical federation that supports HTTPS communication, high-performance privacy set intersection, and efficient communication compression.

Contributors

Thanks goes to these wonderful people:

Zhang Qi, Pan Shaowu, Tan Weicheng, Wu Tiancheng, Xu Yongfei, Zhang Zhaoju, Zhan Zhan, Xu Qiang, Yang Yuan, Li Hongcheng, Tang Cong, Zhang Zhugucheng.

Contributions of any kind are welcome!