Wrappers

Environment wrappers compatible with the RLEnv protocol.

Episode Wrapper

class nnx_ppo.wrappers.episode_wrapper.EpisodeWrapper(env, max_len)[source]

Bases: object

__init__(env, max_len)[source]
step(state, action)[source]
reset(rng)[source]
property observation_size
property action_size

Reward Scaling Wrapper

class nnx_ppo.wrappers.reward_scaling_wrapper.RewardScalingWrapper(env, reward_scale)[source]

Bases: object

__init__(env, reward_scale)[source]
reset(rng)[source]
step(state, action)[source]
property observation_size: Any
property action_size: int