This project is a prototype to demonstrate personalization in federated learning (FedPer) combined with communication-efficient top-k gradient sparsification.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results