Skip to main content

LearnerTransitionBatch

Type Alias LearnerTransitionBatch 

pub type LearnerTransitionBatch<P> = TransitionBatch<<P as Policy>::Observation, <P as Policy>::Action>;
Expand description

Batched transitions for a PolicyLearner.

Aliased Type§

pub struct LearnerTransitionBatch<P> {
    pub states: <P as Policy>::Observation,
    pub next_states: <P as Policy>::Observation,
    pub actions: <P as Policy>::Action,
    pub rewards: Tensor<2>,
    pub dones: Tensor<2>,
}

Fields§

§states: <P as Policy>::Observation

Batched initial states.

§next_states: <P as Policy>::Observation

Batched resulting states.

§actions: <P as Policy>::Action

Batched actions.

§rewards: Tensor<2>

Batched rewards.

§dones: Tensor<2>

Batched flags for terminal states.