NVIDIA FLARE

NVIDIA FLARE (NVIDIA Federated Learning Application Runtime Environment) is a domain-agnostic, open-source, extensible SDK that allows researchers and data scientists to adapt existing ML/DL workflows to a federated paradigm. It enables platform developers to build a secure, privacy-preserving offering for a distributed multi-party collaboration.

Features

FLARE is built on a componentized architecture that allows you to take federated learning workloads from research and simulation to real-world production deployment.

Application Features

Support both deep learning and traditional machine learning algorithms (eg. PyTorch, TensorFlow, Scikit-learn, XGBoost etc.)
Support horizontal and vertical federated learning
Built-in Federated Learning algorithms (e.g., FedAvg, FedProx, FedOpt, Scaffold, Ditto, etc.)
Support multiple server and client-controlled training workflows (e.g., scatter & gather, cyclic) and validation workflows (global model evaluation, cross-site validation)
Support both data analytics (federated statistics) and machine learning lifecycle management
Privacy preservation with differential privacy, homomorphic encryption, private set intersection (PSI)

From Simulation to Real-World

FLARE Client API to transition seamlessly from ML/DL to FL with minimal code changes
Simulator and POC mode for rapid development and prototyping
Fully customizable and extensible components with modular design
Deployment on cloud and on-premise
Dashboard for project management and deployment
Security enforcement through federated authorization and privacy policy
Built-in support for system resiliency and fault tolerance