This repository contains a Federated Domain Generalization (FDG) framework, a machine learning approach that aims to train models capable of generalizing across unseen domains without requiring centralized data collection. It enables clients to collaboratively learn a domain-agnostic model while preserving data privacy.