hanxxyu/pdfgpt-plugin

This repository hosts a Wenxinyiyan plugin that enables the uploading of a PDF URL. The plugin extracts information from the PDF, to enhance the model's responses. It allows for seamless integration of PDF documents into conversations, expanding the model's knowledge and improving its ability to handle complex queries.

Running on localhost

• create conda env

cd backend   # cd root file
conda create -n myenv python=3.9
conda activate myenv

• installing the dependencies

pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple   

• start to deploy web application on port 3333

python plugins.py

Reference

Q1: how to download the Universal Sentence Encoder?

A: click to pdf repo to figure out how to solve this problem

Q2: Is this a RAG application? how to learn the procedure of RAG application?

A:

1. refer to RAG procedure to understand how it works
2. refer to RAG demo to understand how to code
3. refer to RAG improvements to know how to improve a RAG application