This section will lay the groundwork for future coding sections. First, create the following folders for file storage:

DCGAN
├── data
│   └── README.md
├── docker
│   ├── build.sh
│   ├── clean.sh
│   └── Dockerfile
├── README.md
├── scripts

Make sure that you create a DCGAN directory in the $HOME directory in your Ubuntu installation. Many of the scripts we will develop are going to rely on this installation location when mapping volumes using Docker's commands. However, if you are comfortable with changing the mapped volumes in the run command, you won't need to install it in $HOME.

In machine learning (ML), algorithm development is driven by the different datasets that are used for training. For this chapter, we are going to introduce the LSUN dataset. This section will describe how you can download the dataset for this chapter and then format it for the learning step.

In this section, we will cover the following topics:

• Setting up your environment:
  • Dockerfile
  • Build script
  • Clean script
• Collecting LSUN data