Deep learning model creation with .aiff files in python

Question

How can I create a deep learning model using previously recorded .aiff files to train?

I’m trying to create a voice-activated goal horn based on a similar program I found