Intrinisic Gradient Compression for Federated Learning

Luke Melas-Kyriazi, Franklyn Wang

Federated learning is a rapidly-growing area of research which enables a large number of clients to jointly train a machine learning model on privately-held data. One of the larges…