rlkit/torch/sac/sac.py [169:186]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
            self.eval_statistics['QF1 Loss'] = np.mean(ptu.get_numpy(qf1_loss))
            self.eval_statistics['QF2 Loss'] = np.mean(ptu.get_numpy(qf2_loss))
            self.eval_statistics['Policy Loss'] = np.mean(ptu.get_numpy(
                policy_loss
            ))
            self.eval_statistics.update(create_stats_ordered_dict(
                'Q1 Predictions',
                ptu.get_numpy(q1_pred),
            ))
            self.eval_statistics.update(create_stats_ordered_dict(
                'Q2 Predictions',
                ptu.get_numpy(q2_pred),
            ))
            self.eval_statistics.update(create_stats_ordered_dict(
                'Q Targets',
                ptu.get_numpy(q_target),
            ))
            self.eval_statistics.update(create_stats_ordered_dict(
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



rlkit/torch/td3/td3.py [141:158]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
            self.eval_statistics['QF1 Loss'] = np.mean(ptu.get_numpy(qf1_loss))
            self.eval_statistics['QF2 Loss'] = np.mean(ptu.get_numpy(qf2_loss))
            self.eval_statistics['Policy Loss'] = np.mean(ptu.get_numpy(
                policy_loss
            ))
            self.eval_statistics.update(create_stats_ordered_dict(
                'Q1 Predictions',
                ptu.get_numpy(q1_pred),
            ))
            self.eval_statistics.update(create_stats_ordered_dict(
                'Q2 Predictions',
                ptu.get_numpy(q2_pred),
            ))
            self.eval_statistics.update(create_stats_ordered_dict(
                'Q Targets',
                ptu.get_numpy(q_target),
            ))
            self.eval_statistics.update(create_stats_ordered_dict(
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -