Deep model-based reinforcement implementation for CartPole (w. actor-critic) and LunarLander (w. DQN)
Requirements:
- Python 3.x
- Tensorflow
- Numpy
- Gym
- Pickle
- Matplotlib
To train one agent with the model-based version together with actor-critic in CartPole-v0, run the main.py script.
To train one agent with the model-based version together with DQN in LunarLander-v2, run the main_lunar.py script.