The learning continues from where it stopped if a state is found.