大语言模型架构下的智能写作系统设计与实现  

Fine Tuning of Large Language Models and Artificial Intelligence Writing System

作　　者：王建平 WANG Jianping(China Internet Information Center,Beijing 100089,China)

机构地区：[1]中国互联网新闻中心,北京100089

出　　处：《信息与电脑》2023年第22期130-132,共3页Information & Computer

摘　　要：大语言模型(Large Language Models,LLMs)是一种基于深度学习的自然语言处理(Natural Language Processing,NLP)模型,用于理解、生成和处理自然语言文本,逐渐被应用于不同领域。本文通过对LLMs的技术场景探索,利用LLMs对自然语言的语法、语义和上下文信息的理解与生成自然语言的能力,构建改进媒体传统机器写稿模式的智能写作系统。Large Language Models(LLMs)are deep learning-based natural language processing(NLP)models used to comprehend,generate,and process natural language texts through autonomous learning,gradually being applied in different industries.This paper explores the technical scenarios of the large language model,utilizing the technology of the large language model to understand the grammar,semantics,and contextual information of natural language,and to explore the ability to generate natural language,in order to improve the traditional machine writing capabilities in media.

关 键 词：大语言模型 智能写作 新闻发布会 

分 类 号：G212.2[文化科学—新闻学]