e4b6611cb73ef7658f028831be1aa6bd85ecbed0,src/garage/tf/policies/categorical_mlp_policy.py,CategoricalMLPPolicy,get_action,#CategoricalMLPPolicy#Any#,116

Before Change

            dict(numpy.ndarray): Distribution parameters.

        
        sample, prob = self._f_prob(np.expand_dims([observation], 1))
        return np.squeeze(sample[0]), dict(prob=np.squeeze(prob, axis=1)[0])

    def get_actions(self, observations):
        Return multiple actions.

After Change

        
        actions, agent_infos = self.get_actions([observation])
        return actions, {k: v[0] for k, v in agent_infos.items()}

    def get_actions(self, observations):
        Return multiple actions.

Instances