Workload-aware Automatic Parallelization for Multi-GPU DNN Training.
Sungho ShinYoungmin JoJungwook ChoiSwagath VenkataramaniVijayalakshmi SrinivasanWonyong SungPublished in: CoRR (2018)
Keyphrases
- training process
- parallel processing
- training set
- parallel implementation
- real time
- training algorithm
- neural network
- training data
- training phase
- response time
- online learning
- semi automatic
- computer software
- parallel computing
- serious games
- fully automatic
- training examples
- training samples
- data driven
- supervised learning
- feature space