$γ$-FedHT: Stepsize-Aware Hard-Threshold Gradient Compression in Federated Learning

Open in new window