Article 2 Understanding Low-Rank Adaptation (LoRA): A Revolution in Fine-Tuning Large Language Models