Run RL wtih openai baselines

gistfile1.txt

import baselines.common.cmd_util as butils
import baselines.ppo2.ppo2 as ppo
import gym
import bc_gym_planning_env

if __name__ == '__main__':
    # https://github.com/openai/baselines
    env = butils.make_vec_env(
        env_id='EgoCostmapAsImgRandomTurnRoboPlanning-v0',
        env_type='robo_planning',
        num_env=1,
        seed=0,
        flatten_dict_observations=True
    )

    env.reset()
    env.step(env.action_space.sample())

    model = ppo.learn(
        network='mlp',
        env=env,
        total_timesteps=20
    )