NVIDIA FLARE™ (NVIDIA Federated Learning Application Runtime Environment) is a domain-agnostic, open-source, and extensible SDK for Federated Learning. It allows researchers and data scientists to adapt existing ML/DL workflow to a federated paradigm and enables platform developers to build a secure, privacy-preserving offering for a distributed multi-party collaboration.
Privacy-Preserving for Multi-Party Collaboration
Develop and validate more accurate and generalizable AI models from diverse data sources while mitigating the risk of compromising data security and privacy with included privacy-preserving algorithms and workflow strategies.
Accelerate AI Research
Allows researchers and data scientists to adapt existing ML/DL workflow (PyTorch, RAPIDS, Nemo, TensorFlow) to a federated learning paradigm.
General purpose, domain-agnostic federated learning SDK that aims to create an ecosystem of developers, researchers, and data scientists.
What is Federated Learning?
Distributed Multi-Party Collaboration
Federated learning is a way to develop and validate more accurate and generalizable AI models from diverse data sources by mitigating the risk of compromising data security or privacy. It enables AI models to be built with a consortium of data providers without the data ever leaving the individual site.
NVIDIA FLARE provides privacy-preserving algorithms that ensure each change to the global model stays hidden and prevent the server from reverse-engineering the submitted weights and discovering any training data.
Training and Evaluation Workflows
Built-in workflow paradigms use local and decentralized data to keep models relevant at the edge, including learning algorithms for FedAvg, FedOpt, and FedProx.
Extensible Management Tools
Management tools help secure provisioning using SSL certifications, orchestration through an admin console, and monitoring of federated learning experiments using TensorBoard for visualization.
Supports Popular ML/DL Frameworks
Flexible in design, the SDK can be used with PyTorch, Tensorflow, and even Numpy, which allows for integrating federated learning into your current workflow.
Its extensive and open-source API enables researchers to develop new federated workflow strategies, innovative learning, and privacy-preserving algorithms.
Reusable Building Blocks
NVIDIA FLARE provides an easy way to perform federated learning experiments by utilizing the reusable building blocks and example walkthroughs.