!!python/object/apply:collections.OrderedDict
- - - batch_size
    - 512
  - - buffer_size
    - 1000000
  - - gamma
    - 0.98
  - - learning_rate
    - 0.0018514039303149058
  - - learning_starts
    - 1000
  - - n_timesteps
    - 1000000.0
  - - policy
    - MlpPolicy
  - - policy_kwargs
    - dict(net_arch=[256, 256], log_std_init=-2.2692589009754176)
  - - tau
    - 0.05
  - - train_freq
    - 64