Implementation of continuous control experiments from "Calibrated Model-Based Deep Reinforcement Learning" (Malik et. al, 2019).
To run PETS model: python -env cartpole -k 1 -calibrate
To run PETS++: python -env cartpole -k 3
tensorflow<2.4.0>, tensorflow_probability<0.12.1>, gym<0.18.0>, tf_agents<0.7.1>, mujoco_py<2.0.2.9>, numpy<1.19.2>, sklearn<0.23.0>, json<2.0.9>, argparse<1.1>,