Reinforcement learning in simulation. Uses Phong, a Pong simulator.
Requires Tensorflow, numpy, and opencv
Run training on Phong with the command (this will use all default values)
python reinforcing_pong.py
Ctl-C while training will enter a display mode. Ctl-C while in this display mode will terminate the training.
'--help' will show you values you can change. Expect this to take a long time to train well.