general_cfg:
  algo_name: SAC
  collect_traj: false
  device: cpu
  env_name: gym
  interact_summary_fre: 1
  load_checkpoint: false
  load_model_step: best
  load_path: Train_CartPole-v1_SAC_20230618-162702
  max_episode: 200
  max_step: 200
  mode: train
  model_save_fre: 100
  model_summary_fre: 1
  mp_backend: single
  n_learners: 1
  n_workers: 2
  online_eval: true
  online_eval_episode: 10
  seed: 1
  share_buffer: true
algo_cfg:
  action_type: continuous
  actor_layers:
  - activation: relu
    layer_size:
    - 256
    layer_type: linear
  - activation: relu
    layer_size:
    - 256
    layer_type: linear
  actor_lr: 0.0003
  alpha: 0.1
  alpha_lr: 0.0001
  automatic_entropy_tuning: false
  batch_size: 64
  buffer_size: 1000000
  buffer_type: REPLAY_QUE
  critic1_lr: 0.001
  critic2_lr: 0.001
  critic_layers:
  - activation: relu
    layer_size:
    - 256
    layer_type: linear
  - activation: relu
    layer_size:
    - 256
    layer_type: linear
  epsilon_decay: 500
  epsilon_end: 0.01
  epsilon_start: 0.95
  gamma: 0.95
  hidden_dim: 64
  independ_actor: true
  lr: 0.0001
  min_policy: 0
  n_epochs: 1
  n_steps_per_learn: 1
  share_optimizer: false
  start_steps: 10000
  target_update: 1
  target_update_fre: 1
  tau: 0.005
  value_layers:
  - activation: relu
    layer_size:
    - 256
    layer_type: linear
  - activation: relu
    layer_size:
    - 256
    layer_type: linear
env_cfg:
  id: Pendulum-v1
  ignore_params:
  - wrapper
  - ignore_params
  render_mode: null
  wrapper: null