DNSGD is a decentralized normalized stochastic gradient method for (L0,L1)-smooth nonconvex optimization, with complexity bounds that match standard smooth decentralized results when L1=0.
Stochastic gradient push for distributed deep learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
math.OC 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Decentralized Stochastic Nonconvex Optimization under the $(L_0,L_1)$-Smoothness
DNSGD is a decentralized normalized stochastic gradient method for (L0,L1)-smooth nonconvex optimization, with complexity bounds that match standard smooth decentralized results when L1=0.