general_cfg:
  algo_name: A2C
  collect_traj: false
  device: cpu
  env_name: gym
  interact_summary_fre: 1
  load_checkpoint: false
  load_model_step: best
  load_path: Train_single_CartPole-v1_DQN_20230515-211721
  max_episode: 500
  max_step: 200
  mode: train
  model_save_fre: 10
  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_hidden_dim: 256
  actor_layers:
  - activation: relu
    layer_size:
    - 256
    layer_type: linear
  actor_lr: 0.0001
  batch_size: 256
  buffer_type: ONPOLICY_QUE
  critic_hidden_dim: 256
  critic_layers:
  - activation: relu
    layer_size:
    - 256
    layer_type: linear
  critic_loss_coef: 0.5
  critic_lr: 0.005
  entropy_coef: 0.01
  gamma: 0.9
  independ_actor: true
  k_epochs: 4
  lr: 0.0001
  min_policy: 0
  n_steps_per_learn: 1
  sgd_batch_size: 32
  share_optimizer: false
env_cfg:
  id: Pendulum-v1
  ignore_params:
  - wrapper
  - ignore_params
  render_mode: null
  wrapper: null