fb5faeee9ef1ddd3cf40c1f11e2ef92dfaf9a87a,rlkit/samplers/data_collector/step_collector.py,GoalConditionedStepCollector,collect_new_step,#GoalConditionedStepCollector#Any#Any#Any#,60

Before Change

            self._obs[self._desired_goal_key],
        ))
        action, agent_info = self._policy.get_action(new_obs)
        next_ob, reward, terminal, env_info = (
            self._env.step(action)
        )
        terminal = np.array([terminal])
        reward = np.array([reward])
        // store path obs
        self._current_path_builder.add_all(

After Change

            num_steps,
            discard_incomplete_paths,
    ):
        for _ in range(num_steps):
            self.collect_one_step(max_path_length, discard_incomplete_paths)

    def collect_one_step(
            self,
            max_path_length,
            discard_incomplete_paths,

Instances