OakEmu-RL v0.1.0

Appearance

PPOCompletedEpisode ​

python
PPOCompletedEpisode(env_id: int, outcome: EpisodeOutcome, episode_reward: float, episode_length: int, episode_data: dict[str, float], behaviour_kind: TrajectoryBehaviourKind, step: int) -> None

A PPO actor episode that finished while processing one synchronous step.

Source

Values ​

behaviour_kind ​

python
behaviour_kind: TrajectoryBehaviourKind

Source

env_id ​

python
env_id: int

Source

episode_data ​

python
episode_data: dict[str, float]

Source

episode_length ​

python
episode_length: int

Source

episode_reward ​

python
episode_reward: float

Source

outcome ​

python
outcome: EpisodeOutcome

Source

step ​

python
step: int

Source