Using REINFORCE to learn Gym environments -- LunarLander-v2
and CartPole-v0
.
Read the write-up.
Run the program with REPLACE_MODEL
set to False
. If there exists a directory named model
the program will use it to play the environment.
Set RENDER
to True
to enable video.
If model
directory does not exist then the program trains a new model.
If model
directory exists and you wish to train a new model, run the program with REPLACE_MODEL
set to True
.