!!python/object/apply:collections.OrderedDict
-
  - - env
    - InvertedPendulum-v5
  - - algo
    - Hybrid-HD-PPO (HD actor + MLP critic, gradient-adaptive FPE)
  - - teacher_D
    - 512
  - - pruned_D
    - 64
  - - beta_base
    - 0.7071
  - - timesteps_per_stage
    - 1000000
  - - seed
    - 123