722552e32c4c218b00abcdb1464dd2dbcbfa0f2f,src/garage/np/algos/cma_es.py,CMAES,train,#CMAES#Any#,66

Before Change

        self.policy.set_param_values(self._cur_params)
        self._all_returns = []

        return super().train(runner)

    def train_once(self, itr, paths):
        Perform one step of policy optimization given one batch of samples.

After Change

        last_return = None

        for _ in runner.step_epochs():
            for _ in range(self._n_samples):
                runner.step_path = runner.obtain_samples(runner.step_itr)
                last_return = self.train_once(runner.step_itr,
                                              runner.step_path)
                runner.step_itr += 1

        return last_return

    def train_once(self, itr, paths):
        Perform one step of policy optimization given one batch of samples.

Instances