pendulum-ddpg-fake-gpus:
    env: Pendulum-v1
    run: DDPG
    stop:
        episode_reward_mean: -1000
        timesteps_total: 40000
    config:
        # Works for both torch and tf.
        seed: 42
        framework: tf
        actor_hiddens: [64, 64]
        critic_hiddens: [64, 64]
        n_step: 1
        model: {}
        gamma: 0.99
        exploration_config:
            final_scale: 0.02
        min_sample_timesteps_per_reporting: 600
        replay_buffer_config:
          capacity: 10000
        clip_rewards: false
        use_huber: true
        learning_starts: 500
        train_batch_size: 64
        num_workers: 0
        worker_side_prioritization: false
        actor_lr: 0.0001
        critic_lr: 0.0001

        # Fake 2 GPUs.
        num_gpus: 2
        _fake_gpus: true