FATE (Federated AI Technology Enabler) is the world's first industrial grade federated learning open source framework to enable enterprises and institutions to collaborate on data while protecting data security and privacy. It implements secure computation protocols based on homomorphic encryption and multi-party computation (MPC). Supporting various federated learning scenarios, FATE now provides a host of federated learning algorithms, including logistic regression, tree-based algorithms, deep learning and transfer learning.
FATE is an open source project hosted by Linux Foundation. The Technical Charter sets forth the responsibilities and procedures for technical contribution to, and oversight of, the FATE (“Federated AI Technology Enabler”) Project.
https://fate.readthedocs.io/en/latest
FATE can be deployed on a single node or on multiple nodes. Choose the deployment approach which matches your environment. Release version can be downloaded here.
- Deploying FATE on a single node via PyPI, pre-built docker images or installers. It is for simple testing purposes. Refer to this guide.
Deploying FATE to multiple nodes to achieve scalability, reliability and manageability.
- Cluster deployment by CLI: Using CLI to deploy a FATE cluster.
- Training Demo with Only FATE Installed From Pypi
- Training Demo with Both FATE AND FATE-Flow Installed From Pypi
- Architecture: Building Unified and Standardized API for Heterogeneous Computing Engines Interconnection
- FATE Algorithm Components: Building Standardized Algorithm Components for different Scheduling Engines
- OSX (Open Site Exchange): Building Open Platform for Cross-Site Communication Interconnection
- FATE-Flow: Building Open and Standardized Scheduling Platform for Scheduling Interconnection
- PipeLine Design: Building Scalable Federated DSL for Application Layer Interconnection And Providing Tools For Fast Federated Modeling
- RoadMap
- Paper & Conference
- KubeFATE: An operational tool for the FATE platform using cloud native technologies such as containers and Kubernetes.
- FATE-Flow: A multi-party secure task scheduling platform for federated learning pipeline.
- FATE-Board: A suite of visualization tools to explore and understand federated models easily and effectively.
- FATE-Serving: A high-performance and production-ready serving system for federated learning models.
- FATE-Cloud: An infrastructure for building and managing industrial-grade federated learning cloud services.
- EggRoll: A simple high-performance computing framework for (federated) machine learning.
- AnsibleFATE: A tool to optimize and automate the configuration and deployment operations via Ansible.
- FATE-Builder: A tool to build package and docker image for FATE and KubeFATE.
- FATE-Client: A tool to enable fast federated modeling tasks for FATE.
- FATE-Test: An automated testing tool for FATE, including tests and benchmark comparisons.
- FATE-LLM : A framework to support federated learning for large language models(LLMs).
FATE-Community contains all the documents about how the community members cooperate with each other.
- GOVERNANCE.md documents the governance model of the project.
- Minutes of working meetings
- Development Process Guidelines
- Security Release Process
FATE is an inclusive and open community. We welcome developers who are interested in making FATE better! Contributions of all kinds are welcome. Please refer to the general contributing guideline of all FATE projects and the contributing guideline of each repository.
Join the FATE user mailing list, and stay connected with the community and learn about the latest news and information of the FATE project. Discussion and feedback on the FATE project are welcome.
File bugs and features requests via the GitHub issues. If you need help, ask your questions via the mailing list.
Maintainers: FedAI-maintainers @ groups.io
Security Response Committee: FATE-security @ groups.io
Follow us on twitter @FATEFedAI