the number of patches per bag

[wuxchcandoit]

Hi! Thanks for your great work.
After running deepzoom_tiler.py by the command 'python deepzoom_tiler.py -m 1 -b 20 -d Camelyon16 -v tif', I actually get 11,621 patches per bag. But the paper mentions that 'The dataset consists of 271 training images and 129 testing images, which yield roughly 3.2 million patches at 20× magnification and 0.25 million patches at 5× magnification with on average about 8,000 and 625 patches per bag'. I wonder how to get about 8,000 patches per bag. Are there any problems?