Skip to content

AldeaTeam/stock-chat-gpt

Repository files navigation

The journey to a private stock chatbot , with multiple LLMs options, local deployment option, and web search function🔥🚀

This project is heavily inspired by many latest projects in the open-source community with an emphasis on Robby-chatbot for this initial implementation.

Local setup

Win and Mac systems are tested.

  • Clone the repository:
git clone https://github.com/AldeaTeam/stock-chat-gpt.git
  • Navigate to the project directory:
cd stock-chat-gpt
  • Setup a virtual environment:
conda create -n stockchat python== 3.10
  • Install the required dependencies in the virtual environment and tsinghua mirror is strongly recommended!
pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple
  • Launch the chat service locally:
streamlit run src/Home.py

On cloud

AutoDL service is tested and the image will be provided:

  • First input in the new terminal:
source activate 

Core functions and future devplan

  • LLM chat
    • OpenAI ✅
    • Local LLMs like GPT4all, TigerBot...
  • Local knowledge base, vector storage and query
    • Csv, pdf, txt file are supported ✅
    • Graph date to be supported
  • External web search
  • Data visualization and exploration
  • One-click installation
  • And more 😄.
  • Any comments or suggestions will be appreciated.

A record of our efforts and experience.

  • ref/stockchat-brainstorm-plan shows our brainstorming process.
  • ref/linux-deploy-guide.pdf shows the mistakes and how-to-fix guide for linux deployment of LangChain-ChatGLM project.
  • A list of some key open-source projects from which we've gained helpful inspirations are listed below. Many thanks to those creative contributors. 👏🏻👏🏻

记“可私有化部署、切换不同LLM的股票聊天机器人”的学习、开发历程 20230614😂😄

项目最初目标是参考Finchat等应用的功能搭建具备本地PDF总结问答、web搜索总结等功能的股票分析助手,可选用多种LLM模型或商业API。

在社区提供的课程学习基础上,在社区小伙伴的❤️献言献策❤鼓励中以及特别指导 - 社恐杨老师的因材施教、渐进式指导下,(完全没有LLM或编程/产品开发经验的)本小组成员们先后观摩学习了不同的开源项目,从中汲取思路并反复尝试,最后放弃了开源社区比较成熟的LangChain-ChatGLM(虽然基本契合项目开发目标,但环境依赖包略多、配置容易冲突😂),转而专注通过少数框架和工具包,完成核心功能的实现。

核心功能和未来开发计划

  • LLM 对话
    • 商业API 如 openAI ✅
    • 开源本地部署的LLM 如 GPT4all, TigerBot...
  • 本地知识库导入、向量化存储和查询
    • 支持 CSV, PDF, TXT 格式文件✅
    • 支持图数据
  • 外部搜索功能
  • 数据可视化展示和分析模块
  • 一键安装包
  • 待补充,欢迎任何建议和意见~~

参考项目

致谢

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published