The most important part of the lesson after building a model is training! How do you train this beautiful yet simple architecture you have just developed, you might ask? Quite simply, now that we have laid the appropriate framework to do so, the key part is to understand how to run all of these tools that we have developed and then understand the output we are getting from the model.
This is the moment of truth—have you completed all of the previous recipes up until this point? If not, go back and work on them. Your directory should look like the following, minus the items in the data folder if you haven't run the script yet:
full-gan/ ├── data │ ├── Discriminator_Model.png │ ├── GAN_Model.png │ ├── Generator_Model.png │ ├── sample_0.png │ ├── sample_1000.png ├── discriminator.py ├── Dockerfile ├── gan.py ├── generator.py ├── README.md ├── run.py ├── run.sh └── train.py
It's important to get every one of these pieces...