NVIDIA/NVFlare

NVIDIA Federated Learning Application Runtime Environment

What it solves

NVFlare provides a way for researchers and data scientists to train machine learning models across multiple parties without sharing their raw data. It solves the problem of data privacy and security in distributed collaborations, allowing organizations to build privacy-preserving ML workflows.

How it works

It is a domain-agnostic Python SDK that adapts existing ML/DL workflows into a federated paradigm. It supports both horizontal and vertical federated learning and provides built-in algorithms like FedAvg and FedProx. The system uses a componentized architecture that allows users to move from simulation and prototyping (via a Simulator and POC mode) to real-world cloud or on-premise deployment.

Who it’s for

This tool is designed for researchers, data scientists, and platform developers who need to implement secure, multi-party collaboration for machine learning.

Highlights

  • Broad Framework Support: Compatible with PyTorch, TensorFlow, scikit-learn, and XGBoost.
  • Privacy Tools: Includes differential privacy, homomorphic encryption, and private set intersection (PSI).
  • Flexible Workflows: Supports various training (scatter & gather, cyclic) and validation workflows.
  • Production Ready: Features a management dashboard, system resiliency, fault tolerance, and federated authorization.

Related

  • Project
    apple/pfl-research

    A Python framework for simulating private federated learning experiments, supporting multiple ML frameworks and differential privacy mechanisms.

  • Project
    flwrlabs/flower

    A framework for building federated AI systems that enables decentralized model training across various ML frameworks and devices.

  • Project
    google-parfait/tensorflow-federated

    An open-source framework for machine learning and computations on decentralized data, enabling global model training while keeping training data local to clients.

  • Project
    mostly-ai/mostlyai

    A Python SDK for creating high-fidelity, privacy-safe synthetic tabular and language data using local or remote compute resources.

  • Project
    transformerlab/transformerlab-app

    An open-source machine learning platform that unifies training, fine-tuning, inference, and evaluation into a single interface for individuals and research labs.