1 problem

Let $f_i$ denote the local loss function associated with client $i$. Consider decentralized learning settings in which the algorithms either do not suffer from, or actively elimina…