pendulum-ddpg-fake-gpus:
    env: Pendulum-v1
    run: DDPG
    stop:
        episode_reward_mean: -700
        timesteps_total: 100000
    config:
        # Works for both torch and tf.
        framework: tf
        actor_hiddens: [64, 64]
        critic_hiddens: [64, 64]
        n_step: 1
        model: {}
        gamma: 0.99
        exploration_config:
            final_scale: 0.02
        timesteps_per_iteration: 600
        buffer_size: 10000
        clip_rewards: false
        use_huber: true
        learning_starts: 500
        train_batch_size: 64
        num_workers: 0
        worker_side_prioritization: False

        # Fake 2 GPUs.
        num_gpus: 2
        _fake_gpus: true