Continual Learning with Low Rank Adaptation