| general_cfg: | |
| algo_name: TD3_BC | |
| device: cuda | |
| env_name: gym | |
| eval_eps: 5 | |
| eval_per_episode: 1 | |
| load_checkpoint: false | |
| load_path: Train_CartPole-v1_DQN_20221026-054757 | |
| max_steps: 200 | |
| mode: train | |
| mp_backend: mp | |
| new_step_api: true | |
| render: false | |
| render_mode: human | |
| save_fig: true | |
| seed: 1 | |
| show_fig: false | |
| test_eps: 10 | |
| train_eps: 1 | |
| wrapper: null | |
| algo_cfg: | |
| actor_hidden_dim: 256 | |
| actor_lr: 0.0003 | |
| alpha: 5 | |
| batch_size: 100 | |
| buffer_size: 1000000 | |
| critic_hidden_dim: 256 | |
| critic_lr: 0.0003 | |
| expert_path: tasks/Collect_gym_TD3_20230416-111040/traj/traj.pkl | |
| expl_noise: 0.1 | |
| explore_steps: 1000 | |
| gamma: 0.99 | |
| lmbda: 1 | |
| noise_clip: 0.5 | |
| normalize: false | |
| policy_freq: 2 | |
| policy_noise: 0.2 | |
| tau: 0.005 | |
| train_iterations: 1500 | |
| env_cfg: | |
| id: Pendulum-v1 | |
| new_step_api: true | |
| render_mode: null | |