Analyzing the impact of the MPI allreduce in distributed training of convolutional neural networks.
Adrián CastellóMar CatalánManuel F. DolzEnrique S. Quintana-OrtíJosé DuatoPublished in: Computing (2023)
Keyphrases
- convolutional neural networks
- distributed systems
- distributed environment
- cooperative
- supervised learning
- general purpose
- training phase
- test set
- parallel implementation
- convolutional network
- multi agent
- training set
- lightweight
- distributed data
- computer networks
- training examples
- feature extraction
- artificial neural networks
- multiscale