Abstract: Federated learning uses the concept of decentralized training of n number of local clients for a small number of epochs say 2-5, and then averaging the learned weights of all local clients, ...