OakEmu-RL v0.1.0

Appearance

RewardShapingContext ​

python
RewardShapingContext(device: torch.device, metrics: MetricsTracker, observation_shape: tuple[int, ...], observation_dtype: np.dtype[Any], metrics_prefix: Literal['Actor', 'Learner'] = 'Actor') -> None

Runtime context shared with trainer-owned reward shapers.

Source

Values ​

device ​

python
device: torch.device

Source

metrics ​

python
metrics: MetricsTracker

Source

metrics_prefix ​

python
metrics_prefix: Literal['Actor', 'Learner'] = 'Actor'

Source

observation_dtype ​

python
observation_dtype: np.dtype[Any]

Source

observation_shape ​

python
observation_shape: tuple[int, ...]

Source