Chatbot

A web application based on OpenAI API with streamed message communication, Markdown and Latex rendering support. Demo

一个基于OpenAI API的Web聊天机器人，使用流式消息传输，支持Markdown渲染、Latex渲染。Demo

Getting Started

Prerequisite

Please ensure the dependencies below have been installed.

请确保已安装以下环境

Python 3
NodeJs 16
Nginx
Git

Then clone this repository.

然后克隆这个Repo

git clone https://github.com/huanghaozi/chatbot.git

Backend

Install dependencies

Use command below to install dependencies.

使用如下命令安装依赖

cd backend
pip install -r requirements.txt

Run for test

Please modify the API Key in backend/main.py before running.

在运行main.py前，请修改backend/main.py中的API Key

Then run main.py normally for test.

然后正常运行main.py进行测试

python main.py

Run as service

Please create a configure file whose content is like below (Please replace the {content} depend on your environment), a reference for its path is /etc/systemd/system/chatbot-backend.service,

请创建一个配置文件，内容如下(请将{带花括号内容}根据您的环境替换掉)，路径参考/etc/systemd/system/chatbot-backend.service

Attention: Please modify the code to generate secret_key randomly to fixed cause that different worker might use different key if randomly

请注意: 由于worker运行在不同进程上，随机生成key可能导致会话无法保持，请将后端代码中随机生成key的代码app.config['SECRET_KEY'] = os.urandom(24)替换为自定义key

[Unit]
Description=Gunicorn instance to serve chatbot backend
After=network.target

[Service]
WorkingDirectory=/{path-to-project}/backend
Environment="PATH=/{path-to-python}/bin"
ExecStart=/{path-to-python}/bin/gunicorn -w4 -b127.0.0.1:1217 main:app --timeout 300

[Install]
WantedBy=multi-user.target

Then enable and start the service using systemctl command as below.

然后使用systemctl命令启用并打开这个服务，内容如下

systemctl enable chatbot-backend.service
systemctl start chatbot-backend.service

Reverse proxy configuration

Please add a piece of code in Nginx configuration file as below and replace your.domain.com to yours.

请将如下代码添加进Nginx配置文件中，并将your.domain.com换成您自己的域名

server {
    listen 80;
    server_name your.domain.com;
    location / {
         add_header Access-Control-Allow-Origin *;
         add_header Access-Control-Allow-Methods 'GET, POST, OPTIONS';
         add_header Access-Control-Allow-Headers 'DNT,X-Mx-ReqToken,Keep-Alive,User-Agent,X-Requested-With,If-Modified-Since,Cache-Control,Content-Type,Authorization';
         proxy_set_header Connection '';
         proxy_http_version 1.1;
         chunked_transfer_encoding off;
         proxy_buffering off;
         proxy_cache off;
         proxy_pass http://127.0.0.1:1217/;
    }
}

Frontend

Install dependencies

Use command below to install dependencies.

使用如下命令安装依赖

npm install

Or use yarn if you can

有条件也可以用Yarn

yarn install

Generate code for Production

Use command below to generate code for production.

使用如下命令生成生产环境代码

npm run build

Or use yarn if you can

有条件也可以用Yarn

yarn run build

Configure Nginx

Use configuration below to deploy the frontend project. Please replace your.domain.com and {path-to-dist} depend on your environment to get the configuration valid.

使用如下Nginx配置部署前端，请根据您自己的环境修改your.domain.com and {path-to-dist}使配置生效

server {
    listen 80;
    server_name your.domain.com;
    location / {
        root /{path-to-dist};
        index index.html index.htm;
    }
}

Run

Please ensure that your domain DNS configuration is right. And then use the command below to test if Nginx configuration is valid.

请确保您正确配置了域名DNS服务，然后使用如下命令测试Nginx配置文件的有效性

sudo nginx -t

After these steps, you can restart the Nginx service to apply the configuration.

完成以上步骤后，您可以重启Nginx服务以应用配置

sudo service nginx restart