Dataset:
Training samples: 20000 images (10000 per class)
Validation samples: 50000 images (2500 per class)
Testing samples: 12500 unlabelled images
PyTorch ConvNet architecture
(conv_1): Conv2d(3, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(conv_2): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(maxpool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
(drop): Dropout(p=0.2, inplace=False)
(fc1): Linear(in_features=160000, out_features=512, bias=True)
(out): Linear(in_features=512, out_features=2, bias=True)
Training loss: 0.392
Validation loss: 0.40058
Test loss: 0.427
Test accuracy: 0.85