TrialScheduler choosing trials to run

Ray Libraries (Data, Train, Tune, Serve)
1

I use a ResouceChangingScheduler and would like to continue running trials with changed resources before starting new trials. Since trials are automatically paused when allocating new resources, I modified the choose_trial_to_run method of the TrialScheduler class in the following way.

    def choose_trial_to_run(self, tune_controller: "TuneController") -> Optional[Trial]:
      for trial in tune_controller.get_trials():
          if trial.status == Trial.PAUSED:
              return trial   
        return None

However, the trial scheduler still starts new trials instead of re-starting the paused ones. How do I need to modify the method to achieve what I want?