Exact and Linear Convergence for Federated Learning under Arbitrary Client Participation is Attainable

Open in new window