# Codeblock 12
encoder_torch = EncoderTorch()

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