SCAR: Efficient Instruction-Tuning for Large Language Models via Style Consistency-Aware Response Ranking

Open in new window