Argument | Improved algorithm in this paper, ResNet-50, GoogleNet, MobileNetv1, VGG16 |
Learning rate | 0.001 |
optimizer | Adam |
Loss function | Categorical |
Batch size | 8 |
Training times | 15 |