colosseum.agent.actors
The actor components of reinforcement learning agents handle the interactions with the MDPs.
Their main role is to provide a mapping from a BaseMDPModel, which contains the knowledge of the agent regarding the
MDP, to a policy.
1""" 2The actor components of reinforcement learning agents handle the interactions with the MDPs. 3Their main role is to provide a mapping from a `BaseMDPModel`, which contains the knowledge of the agent regarding the 4MDP, to a policy. 5""" 6 7from typing import Union 8 9from colosseum.agent.actors.base import BaseActor 10from colosseum.agent.actors.Q_values_actor import QValuesActor 11from colosseum.agent.actors.random import RandomActor 12 13ACTOR_TYPES = Union[BaseActor, QValuesActor]