Zobrazeno 1 - 1
of 1
pro vyhledávání: '"Hao, Chuzhan"'
Instruction Tuning has the potential to stimulate or enhance specific capabilities of large language models (LLMs). However, achieving the right balance of data is crucial to prevent catastrophic forgetting and interference between tasks. To address
Externí odkaz:
http://arxiv.org/abs/2403.03432