cartpole-sac:
    env: CartPoleContinuousBulletEnv-v0
    run: SAC
    stop:
        episode_reward_mean: 40
        timesteps_total: 100000
    config:
        # Works for both torch and tf.
        framework: tf
        gamma: 0.95
        no_done_at_end: false
        horizon: 200
        soft_horizon: true
        n_step: 3
        prioritized_replay: true
        initial_alpha: 0.2
        learning_starts: 256
        clip_actions: false
        timesteps_per_iteration: 1000
        optimization:
            actor_learning_rate: 0.005
            critic_learning_rate: 0.005
            entropy_learning_rate: 0.0001