(Game library) ▼
You can change the game this AI learns by going to line 74 and changing the gym.make variable to any other game in the Open-AI gym library. Here is a link to all of the games. Some work better then others with this specific learning algorithm. https://gym.openai.com/envs/#atari
(Working enviroment ▼)
This project was built with slightly older syntax from tensorflow 1.14.0-1.15.0 and may not work with some builds of tensorflow 2.0. I reccomend using a docker file or virtual enviroment if you regularly use tensorflow 2.
(repos) ▼
Open-AI gym https://github.com/openai/gym
Numpy https://github.com/numpy/numpy
Tensorflow 1.15.0 (downgrade if necessary) https://github.com/tensorflow/tensorflow