Zobrazeno 1 - 3
of 3
pro vyhledávání: '"Zeng, Yuanhao"'
Although instruction tuning is widely used to adjust behavior in Large Language Models (LLMs), extensive empirical evidence and research indicates that it is primarily a process where the model fits to specific task formats, rather than acquiring new
Externí odkaz:
http://arxiv.org/abs/2408.10841
Large Language Models (LLMs) have excelled in various tasks but perform better in high-resource scenarios, which presents challenges in low-resource scenarios. Data scarcity and the inherent difficulty of adapting LLMs to specific tasks compound the
Externí odkaz:
http://arxiv.org/abs/2404.00914
Autor:
Zhuang, Shiyang 1, 2, 3, Zeng, Yuanhao 4, Lin, Shaojunjie 3, Chen, Xirui 3, Xin, Yishan 5, 6, 7, 8, Li, Hongyan 5, 6, 7, 8, Lin, Yiming 5, 6, 7, 8, Zhang, Chaofan 5, 6, 7, 8, ∗, Lin, Yunzhi 1, 2, 9, ∗∗
Publikováno v:
In iScience 20 December 2024 27(12)
