GradientFlow: Optimizing Network Performance for Large-Scale Distributed DNN Training.
Peng SunYonggang WenRuobing HanWansen FengShengen YanPublished in: IEEE Trans. Big Data (2022)
Keyphrases
- distributed network
- peer to peer
- training process
- computer networks
- communication cost
- communication overhead
- data transfer
- network nodes
- radial basis function network
- massive scale
- neural network structure
- network structure
- camera network
- complex networks
- wireless sensor networks
- training set
- multi agent
- load balance
- information dissemination
- network model
- recurrent networks
- central server
- query routing
- communication networks
- distributed systems
- neural network
- remote sites
- multilayer neural network
- data sets
- distributed control
- fully distributed
- heterogeneous networks
- network management
- training phase
- load balancing
- lightweight
- supervised learning