Rina: Enhancing Ring-AllReduce with In-network Aggregation in Distributed Model Training.
Zixuan ChenXuandong LiuMinglin LiYinfan HuHao MeiHuifeng XingHao WangWanxin ShiSen LiuYang XuPublished in: CoRR (2024)
Keyphrases
- network model
- computational model
- computer networks
- probabilistic model
- management system
- experimental data
- formal model
- distributed network
- cooperative
- prior knowledge
- propagation model
- training process
- network structure
- conceptual model
- communication overhead
- radial basis function network
- neural network
- cooperative game
- distributed environment
- statistical model
- social network analysis
- mathematical model
- peer to peer
- objective function
- training data
- high level