Xin Guo

Latest

• FedSkel: Efficient Federated Learning on Heterogeneous Systems with Skeleton Gradients Update
• Accelerating CNN Training by Pruning Activation Gradients