Optimizing distributed training deployment in heterogeneous GPU clusters.
Xiaodong YiShiwei ZhangZiyue LuoGuoping LongLansong DiaoChuan WuZhen ZhengJun YangWei LinPublished in: CoNEXT (2020)
Keyphrases
- loosely coupled
- distributed information systems
- distributed environment
- distributed systems
- distributed data
- clustering algorithm
- heterogeneous environments
- heterogeneous databases
- heterogeneous data
- cooperative
- multi agent
- mobile agents
- training process
- real time
- heterogeneous networks
- distributed data sources
- distributed heterogeneous
- supervised learning
- distributed computing
- cluster analysis
- training algorithm
- parallel processing
- computer networks
- test set
- training set
- hierarchical clustering
- gpu accelerated
- transparent access
- heterogeneous computing
- highly distributed
- training examples
- online learning
- input data
- general purpose
- web services