because of its file size I can't push train_data.binaryfile and test_data.binaryfile, so even if you clone these, these would not work.
cat_CNN_model.json is model file. and cat_CNN_result.h5 is weight file. cat_recognition_program.py is main program.
cat_CNNver2 is notebook for creating this program, so all code that is created this program is expressed in this notebook. But this is written carelessly, so file path is needed to rewrite according to your environment. And some of file in this note may not be necessary.
As you can see notebook result, accuracy for test data is almost 75%.