4dd96193aaa69485ef183778f053fcfcbef457a2,sac/algos/base.py,RLAlgorithm,_train,#RLAlgorithm#Any#Any#Any#,72

Before Change

                        observation = next_ob
                    gt.stamp("sample")

                    if self._pool.size >= self._min_pool_size:
                        for i in range(self._n_train_repeat):
                            batch = self._pool.random_batch(self._batch_size)
                            self._do_training(iteration, batch)

                    gt.stamp("train")

                self._evaluate(epoch)

After Change

                gt.stamp("eval")

            self.sampler.terminate()

    def _evaluate(self, epoch):
        Perform evaluation for the current policy.

Instances