Finally, to run the run.py code, we need to run create a shell script called run.sh, make it executable, and put the following code inside the script:

#/bin/bash

# Training Step
xhost +
docker run -it 
   --runtime=nvidia 
   --rm 
   -e DISPLAY=$DISPLAY 
   -v /tmp/.X11-unix:/tmp/.X11-unix 
   -v $HOME/simGAN/out:/out 
   -v $HOME/simGAN/src:/src 
 ch7 python3 /src/run.py

To run this script, in the root directory of your simGAN chapter:

sudo ./run.sh

That completes the training section of the simGAN chapter. I encourage you to try other datasets with this architecture including experimenting with different loss functions.