# Codeblock 10
encoder = Encoder()

images = torch.randn(BATCH_SIZE, IN_CHANNELS, IMAGE_SIZE, IMAGE_SIZE)
features = encoder(images)