Struct ActorCriticAgent

pub struct ActorCriticAgent<OS, AS, P: Policy, C> { /* private fields */ }

Actor-crtic agent. Consists of a Policy and a Critic.

Implementations

impl<OS, AS, P: Policy, C> ActorCriticAgent<OS, AS, P, C>

where OS: FeatureSpace, AS: ParameterizedDistributionSpace<Tensor>,

pub fn new<E, PB, CB>(env: &E, config: &ActorCriticConfig<PB, CB>) -> Self

where E: EnvStructure<ObservationSpace = OS, ActionSpace = AS> + ?Sized, E::FeedbackSpace: Space<Element = Reward>, PB: BuildPolicy<Policy = P>, CB: BuildCritic<Critic = C>,

Trait Implementations

impl<OS, AS, P, C> Agent<<OS as Space>::Element, <AS as Space>::Element> for ActorCriticAgent<OS, AS, P, C>

where OS: FeatureSpace + Clone, AS: ParameterizedDistributionSpace<Tensor> + Clone, P: Policy,

type Actor = PolicyActor<OS, AS, <P as AsModule>::Module>

Produces actions for a sequence of environment observations.

fn actor(&self, _: ActorMode) -> Self::Actor

Create a new Actor with the given behaviour mode.

impl<OS, AS, P, C> BatchUpdate<<OS as Space>::Element, <AS as Space>::Element> for ActorCriticAgent<OS, AS, P, C>

where OS: FeatureSpace, OS::Element: 'static, AS: ParameterizedDistributionSpace<Tensor>, AS::Element: 'static, P: Policy, C: Critic,

type Feedback = Reward

Environment feedback type

type HistoryBuffer = VecBuffer<<OS as Space>::Element, <AS as Space>::Element>

fn buffer(&self) -> Self::HistoryBuffer

Create a new history buffer.

fn min_update_size(&self) -> HistoryDataBound

Request a minimum amount of on-policy data for the next batch update.

fn batch_update<'a, I>(&mut self, buffers: I, logger: &mut dyn StatsLogger)

where I: IntoIterator<Item = &'a mut Self::HistoryBuffer>, Self::HistoryBuffer: 'a,

Update the agent from a collection of history buffers.

impl<OS: Clone, AS: Clone, P: Clone + Policy, C: Clone> Clone for ActorCriticAgent<OS, AS, P, C>

fn clone(&self) -> ActorCriticAgent<OS, AS, P, C>

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl<OS: Debug, AS: Debug, P: Debug + Policy, C: Debug> Debug for ActorCriticAgent<OS, AS, P, C>

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<'de, OS, AS, P, C> Deserialize<'de> for ActorCriticAgent<OS, AS, P, C>

where OS: Deserialize<'de>, AS: Deserialize<'de>, P: Deserialize<'de> + Policy, C: Deserialize<'de>,

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl<OS: PartialEq, AS: PartialEq, P: PartialEq + Policy, C: PartialEq> PartialEq for ActorCriticAgent<OS, AS, P, C>

fn eq(&self, other: &ActorCriticAgent<OS, AS, P, C>) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl<OS, AS, P, C> Serialize for ActorCriticAgent<OS, AS, P, C>

where OS: Serialize, AS: Serialize, P: Serialize + Policy, C: Serialize,

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl<OS, AS, P: Policy, C> StructuralPartialEq for ActorCriticAgent<OS, AS, P, C>