q_derived.base_q_derived_soft_policy

Classes

BaseQDerivedSoftPolicy

Helper class that provides a standard way to create an ABC using

Module Contents

class q_derived.base_q_derived_soft_policy.BaseQDerivedSoftPolicy(Q: Any | Mapping, epsilon: float, num_actions: int)

Bases: gridmind.policies.soft.base_soft_policy.BaseSoftPolicy

Helper class that provides a standard way to create an ABC using inheritance.

Q

epsilon

num_actions

action_space = None

abstract update(state, action)

get_epsilon()

set_epsilon(value: float)

abstract decay_epsilon()

_get_random_action(action_mask=None)

get_action(state, action_mask=None)

get_action_prob(state, action, action_mask=None)

get_all_action_probabilities(states, action_mask=None)

get_action_deterministic(state, action_mask=None)

abstract _get_greedy_action(state, action_mask=None)