45496de13b7932d8ec860da06f14d113000803be,ml/rl/training/dqn_trainer.py,DQNTrainer,calculate_q_values,#DQNTrainer#Any#,94

Before Change

        return len(self._actions)

    def calculate_q_values(self, states):
        return self.q_network(states).detach()

    def get_max_q_values(self, states, possible_actions, double_q_learning):
        
        Used in Q-learning update.

After Change

        is_numpy = False
        if isinstance(states, np.ndarray):
            is_numpy = True
            states = torch.tensor(states).type(self.dtype)
        result = self.q_network(states).detach()
        if is_numpy:
            return result.cpu().numpy()
        else:
            return result

    def get_max_q_values(self, states, possible_actions, double_q_learning):

Instances