The Best of Both Worlds: Accurate Global and Personalized Models through Federated Learning with Data-Free Hyper-Knowledge Distillation.

Huancheng Chen, Chianing Wang, Haris Vikalo

Browse the full ICLR paper archive.