FedHF is a Python-based simulator for flexible, heterogeneous, and asynchronous federated learning research. It provides configurable resource models, supports asynchronous protocols, and accelerates experimentation.

Features

  • Loose coupling of client/server for FL simulations
  • Supports asynchronous federated protocols
  • Models heterogeneous device/resource performance
  • Configurable via Python APIs
  • Accelerates prototyping of FL algorithms
  • Apache‑2.0 licensed with documentation

Project Samples

Project Activity

See All Activity >