This is an implementation of PPO algorithm.

Run the following command to start parallelized training:

python experiment.py

One could modify experiment.py to quickly set up different configurations.