Training EfficientNets at Supercomputer Scale: 83% ImageNet Top-1 Accuracy in One Hour.
Arissa WongpanichHieu PhamJames DemmelMingxing TanQuoc V. LeYang YouSameer KumarPublished in: CoRR (2020)
Keyphrases
- high accuracy
- computational cost
- error rate
- linear svm
- test set
- prediction accuracy
- training and testing data
- data sets
- training process
- hierarchical structure
- training set
- multiscale
- supervised learning
- active learning
- feature vectors
- artificial neural networks
- highly accurate
- training algorithm
- floating point
- computational complexity
- classification performances