level_replay/model.py [152:165]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
        dist = self.dist(actor_features)

        if deterministic:
            action = dist.mode()
        else:
            action = dist.sample()

        # action_log_probs = dist.log_probs(action)
        action_log_dist = dist.logits
        dist_entropy = dist.entropy().mean()

        return value, action, action_log_dist, rnn_hxs

    def get_value(self, inputs, rnn_hxs, masks):
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



level_replay/model.py [476:489]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
        dist = self.dist(actor_features)

        if deterministic:
            action = dist.mode()
        else:
            action = dist.sample()

        # action_log_probs = dist.log_probs(action)
        action_log_dist = dist.logits
        dist_entropy = dist.entropy().mean()

        return value, action, action_log_dist, rnn_hxs

    def get_value(self, inputs, rnn_hxs, masks):
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -