general_cfg: algo_name: DDPG device: cuda env_name: Pendulum-v1 eval_eps: 10 eval_per_episode: 5 load_checkpoint: false load_path: Train_CartPole-v1_DQN_20221026-054757 max_steps: 200 mode: train new_step_api: true render: false save_fig: true seed: 10 show_fig: false test_eps: 10 train_eps: 400 wrapper: null algo_cfg: actor_hidden_dim: 256 actor_lr: 0.0001 batch_size: 128 buffer_size: 8000 critic_hidden_dim: 256 critic_lr: 0.001 gamma: 0.99 tau: 0.001