!!python/object/apply:collections.OrderedDict
-
  - - env
    - Pendulum-v1
  - - algo
    - HD-PPO (gradient-adaptive FPE, continuous)
  - - teacher_D
    - 512
  - - pruned_D
    - 32
  - - beta_base
    - 2.5
  - - timesteps_per_stage
    - 1000000
  - - seed
    - 2024