Requirements to install for Genetic Algorithm: gym py4j port_for #Although I have replace the library with import socket instead, because port_for was not working properly opencv-python
then running pip install -e . within the FightingICEV4.40 folder
To run our AI, you have to navigate into FightingICEV4.40 sub-directory:
In both BeginnerAI.py and BegInterAI.py, you have to change the line for self.env = gym.make("Fightinginc...", java_env_path="(The directory to the FightingICEV4.40)")
You will have to manually change the path, so it leads directly to the FightingICEV4.40 sub-directory
For the genetic algorithm implementation:
1.) Run BeginnerAI.py
For the DQN reinforcement algorithm, you will have to take extra steps:
1.) Make sure that your system is compatible with CUDA
2.) Ensure you are the environment is running on python 3.6
3.) Install keras and tensorflow-gpu
4.) Download CUDA toolkit (which includes the driver)
a.) https://docs.nvidia.com/cuda/cuda-installation-guide-microsoft-windows/ for more in-depth details
b.) at the bottom of https://www.tensorflow.org/install/gpu are described extra steps if it is not working
5.) Reset computer
6.) Run BegIterAI.py in the FightingICEV4.40 sub-directory
7.) If tensorflow-gpu can't locate cudart64_100.dll (only if you get an error)
i.) Navigate to CUDAs bin directory (C:/Program Files/ NVIDIA GPU Computing Toolkit/CUDA/v10.2/bin
ii.) Drag the cudart64_100.dll file from the zip file within the root directory of this project
8.) Run BegIterAI.py