algorithms.util.action_selector.BoltzmannActionSelector

class algorithms.util.action_selector.BoltzmannActionSelector(initial_tau: float, tau_decay: bool, tau_decay_coeff: float)[source]

Bases: ActionSelector

Implements the Boltzmann policy.

__init__(initial_tau: float, tau_decay: bool, tau_decay_coeff: float)[source]

Methods

__init__(initial_tau, tau_decay, tau_decay_coeff)

choose(values, step)