Efficient Task Adaptation in Large Language Models via Selective Parameter Optimization

Weijie Wan, Jiangjiang Zhao

Large Language Models (LLMs) have demonstrated excellent performance in general language understanding, generation and other tasks. However, when fine-tuning for specific domain ta…