RAG-in-LangGraph

中文版本

Project Description

This is a RAG (Retrieval Augmented Generation) project based on LangGraph. It aims to reduce the barrier to entry for RAG, making it ready to use.

This project includes retrieval augmentation, allowing users to chat with uploaded PDFs, achieving a chatpdf function. By combining retrieval and generation models, it provides more accurate and information-rich answers.

Achieve RAG augmentation with just a few lines of code:

from dotenv import load_dotenv
load_dotenv(".env")

from rag_tools import RAGTool

rag_tool = RAGTool()
file_path = "README.pdf"
rag_tool.load_pdf(file_path)

message = rag_tool.query("Please design 3 multiple-choice questions to test my understanding of the material.")
print(message.content)

Installation

1. Clone the project locally:

   git clone https://github.com/kuhung/RAG-in-LangGraph.git
   cd RAG-in-LangGraph

2. Install uv (if not already installed):

   curl -LsSf https://astral.sh/uv/install.sh | sh
   # Or install via pipx: pipx install uv

3. Install Python dependencies:

   uv sync

4. Configure environment variables: Copy the .env.example file and rename it to .env, then fill in API keys and other information according to your actual situation.

Usage

Currently, the project includes rag_tools.py for defining RAG-related tool functions. The specific workflow is as follows:

graph TD
    A[User calls query method] --> B[Build state graph]
    B --> C[START]
    C --> D[Retrieval node]
    D -->|Similarity search| E[Vector store]
    E -->|Return top-k documents| D
    D --> F[Generation node]
    F -->|Use context| G[LLM generates answer]
    G --> H[END]
    H --> I[Return answer to user]

Loading

The rag.ipynb file demonstrates the process. Only a few lines of code are needed to call the complete RAG lifecycle:

from dotenv import load_dotenv
load_dotenv(".env")

from rag_tools import RAGTool

rag_tool = RAGTool()
file_path = "README.pdf"
rag_tool.load_pdf(file_path)

message = rag_tool.query("Please design 3 multiple-choice questions to test my understanding of the material.")
print(message.content)

Contact

• author: kuhung
• Email: hi@kuhung.me

License

This project is licensed under the MIT License - see the LICENSE file for details.

---

RAG-in-LangGraph

项目描述

这是一个基于 LangGraph 实现 RAG (Retrieval Augmented Generation) 的项目。它旨在减少关于RAG的上手成本，拿来就能用。

该项目包含检索增强，用户可与上传的PDF进行对话，实现 chatpdf 功能。通过结合检索和生成模型，提供更准确和信息丰富的回答。

只需几行代码，即可调用实现RAG增强

from dotenv import load_dotenv
load_dotenv(".env")

from rag_tools import RAGTool

rag_tool = RAGTool()
file_path = "README.pdf"
rag_tool.load_pdf(file_path)

message = rag_tool.query("帮我设计3个选择题，考验我对资料的掌握情况")
print(message.content)

安装

1. 克隆项目到本地：

   git clone https://github.com/kuhung/RAG-in-LangGraph.git
   cd RAG-in-LangGraph

2. 安装 uv (如果尚未安装)：

   curl -LsSf https://astral.sh/uv/install.sh | sh
   # 或者通过 pipx 安装：pipx install uv

3. 安装 Python 依赖：

   uv sync

4. 配置环境变量： 复制 .env.example 文件并重命名为 .env，然后根据您的实际情况填写 API 密钥等信息。

说明

目前项目包含 rag_tools.py 文件，用于定义 RAG 相关的工具函数。具体工作流程如下：

graph TD
    A[用户调用 query 方法] --> B[构建状态图]
    B --> C[START]
    C --> D[检索节点]
    D -->|相似度搜索| E[向量存储]
    E -->|返回 top-k 文档| D
    D --> F[生成节点]
    F -->|使用上下文| G[LLM 生成回答]
    G --> H[END]
    H --> I[返回答案给用户]

Loading

文件 rag.ipynb为演示流程，只需要下面的几行代码，即可调用完整的RAG生命周期：

from dotenv import load_dotenv
load_dotenv(".env")

from rag_tools import RAGTool

rag_tool = RAGTool()
file_path = "README.pdf"
rag_tool.load_pdf(file_path)

message = rag_tool.query("帮我设计3个选择题，考验我对资料的掌握情况")
print(message.content)

联系方式

• 作者: kuhung
• 邮箱: hi@kuhung.me

许可证

本项目使用 MIT 许可证 - 更多详情请参阅 LICENSE 文件。