tf_agents.bandits.agents.bernoulli_thompson_sampling_agent.BernoulliBanditVariableCollection

A collection of variables used by BernoulliThompsonSamplingAgent.

tf_agents.bandits.agents.bernoulli_thompson_sampling_agent.BernoulliBanditVariableCollection(
    num_actions: int, dtype: tf.DType = tf.float32, name: Optional[Text] = None
)

Args
num_actions	(int) The number of actions.
dtype	The type of the variables. Should be one of tf.float32 or tf.float64.
name	(string) the name of this instance.