Faster Federated Learning with Decaying Number of Local SGD Steps

Open in new window