Abstract: Decentralized stochastic gradient descent (SGD) is a driving engine for decentralized federated learning (DFL). The performance of decentralized SGD is jointly influenced by inter-node ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results