Data is the lifeblood of these algorithms. If you take nothing else away from this book, please learn this lesson. In this recipe, we'll read each of the files in an array, resize them for learning, and save them into an easy-to-access compressed format.
First, let's perform a sanity check on our directory structure to make sure that we have all the right pieces; it should look as follows:
DCGAN ├── data │ └── README.md ├── docker │ ├── build.sh │ ├── clean.sh │ └── Dockerfile ├── README.md ├── scripts │ └── create_data.sh └── src ├── save_to_npy.py
You should notice the new folder, src
, along with a new file, save_to_npy.py
. The following recipe will focus on this Python file and how to run it to save data.