基于LLM+RAG的跑团小助手 #432
SingaCycle
started this conversation in
project
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
1. 背景
个人跑团助手,利用大型语言模型(LLM)和检索增强生成模型(RAG)的组合。作为跑团支持,角色扮演以及小说创作中的个人助手。主要是实现包括关键规则以及背景故事的查询,场景设定和推演,剧情模拟等功能。
2.项目内容
(1) 收集代码/书籍数据,作为微调数据集或者知识库外挂
(2) 设计Prompt模板
(3) 搭建LLM工作流程
(4) 前端展示页面
(5) 基于微调,是LLM的生成风格更符合任务(待定)
(6) 基于知识图谱(小说人物关系,事件流程)+RAG的高效检索(待定)
(7) 生成内容的结构化展示(人物关系之类)(待定)
3.暂定技术栈
(1) 基座模型:InternLM,Qwen的开源7B
(2) 微调框架:XTuner
(3) Chain,Prompt框架:Langchain
(4) RAG框架: Langchain, LLamma Index
(5) 部署: LMDeploy
(6) 展示: streamlit
Beta Was this translation helpful? Give feedback.
All reactions