FedDA: Faster Framework of Local Adaptive Gradient Methods via Restarted Dual Averaging

Open in new window