cartpole-ddppo:
    env: CartPole-v0
    run: DDPPO
    stop:
        episode_reward_mean: 100
        timesteps_total: 100000
    config:
        num_gpus_per_worker: 0