spotDNN: Provisioning Spot Instances for Predictable Distributed DNN Training in the Cloud.
Ruitao ShangFei XuZhuoyan BaiLi ChenZhi ZhouFangming LiuPublished in: IWQoS (2023)
Keyphrases
- cloud computing
- training process
- cloud computing environment
- distributed computing
- distributed systems
- data center
- training algorithm
- training set
- training phase
- lightweight
- distributed data
- randomly generated
- distributed environment
- multi agent
- computing resources
- training samples
- map reduce
- mobile agents
- training data
- virtual machine
- neural network
- communication cost
- trained classifiers
- databases
- telecommunication services
- resource management
- computer networks
- quality of service
- test set
- online learning
- peer to peer
- search space
- data streams
- cooperative
- feature selection