chatgpt dingtalk Download - chatgpt dingtalk Source code download

chatgpt dingtalk

Other source code

v1.1.6 2023-07-26?

Download

ChatGPT DingTalk

? DingTalk integrated ChatGPT robot based on GO language?

Table of contents

Preface
Functional introduction
Prerequisites for use
Usage tutorial
- The first step is to deploy the application
  - docker deployment
  - Binary deployment
- Step 2: Add an application
Highlights
- Private chat with the robot
- Help List
- Switch mode
- Query the balance
- Daily issues
- Chat with built-in propt
- Generate pictures
- Support gpt-4
Local development
Configuration file description
Frequently Asked Questions
Join the group to communicate
grateful
Appreciate
Highlight moments
Star History
Contributor List

Preface

This project can help you integrate GPT robots into DingTalk Group Chat. The current default model is gpt-3.5 , and it supports gpt-4 and gpt-4o-mini . It also supports Azure-OpenAI.

? 注意: The deployment and configuration process are now very mature, and the documents and issue are basically covered. Therefore, no questions about installation, deployment and configuration use of any project will be answered. If you don’t understand at all, you can consider contacting me through email for paid technical support.
? 注意: All the functions of this project are gathered in the usage guide. Please be sure to read them carefully to experience its complete essence.

? Welcome to follow my other open source projects:

Go-Ldap-Admin:? OpenLDAP backend management project based on Go+Vue.
learning-weekly: The weekly content is mainly operation and maintenance technology and Go language peripherals, supplemented by excellent projects on GitHub or other people's excellent experience.
HowToStartOpenSource:? GitHub Open Source Project Maintenance Collaboration Guide.
read-list: The reading is the basis for high-quality content subscription
awesome-github-profile-readme-chinese:? Collection of excellent Chinese area personal homepages

? I also created a project awesome-chatgpt-answer : record those moments of good questions and great answers. Welcome to submit the exquisite conversations you encountered during your interaction with ChatGPT.

⚗️ openai official provides a status page to present the current status of the openAI service. At the same time, if there are any problems, the announcement will be posted on this page. If you feel that it has a problem, you can check it out on this page.

Sponsors

302.AI is a one-stop AI application platform that pays on demand, an open platform, and an open source ecosystem.
Click to register: Get 1PTC (1PTC=1 US dollars, about RMB 7) tokens immediately.
It brings together the latest and most complete AI models and brands, including but not limited to language models, image models, sound models, and video models.
Carry out in-depth application development on the basic model so that novices can use it at zero threshold without learning costs.
Zero monthly fee, all functions are paid on demand, fully open, achieving a real low threshold and high ceiling.
Innovative usage model, separation of management and use, aimed at teams and small and medium-sized enterprises, managed by one person and used by multiple people.
All AI capabilities provide API access, and all applications open source support self-customization (in progress).
A strong development team, launching 2-3 new applications every week, and the platform features are updated daily.

Functional introduction

Help menu: By sending帮助you will see the help list,? View the example
? Private chat: Supports separate private chat with robots (no need to have sex), ? View example
Group chat: Support conversations in the group Aite robot
? Single chat mode: Each conversation is a new conversation, without historical chat context contact
? Chat mode: a dialogue mode with contextual understanding
? Image generation: Generate images by sending content starting with #图片keyword, ? View example
? Role-playing: Support scene mode, triggering the built-in propt template through #周报? View the example
?‍ Frequency limit: By specifying configuration, customize the maximum number of conversations per day for a single user
? Balance query: Query the remaining amount of the current key by sending余额keyword, ? View the example
? Custom API domain name: Through configuration specification, solve the problem that domestic servers cannot directly access openai
? Add a proxy: By configuring the configuration, solve the problem of inaccessibility of domestic servers by injecting a proxy into the application.
? Default mode: Supports custom default chat mode, specified by configuration
Query dialogue: By sending #查对话username:xxx xxx, you can preview it online and download it to the local area.
? Whitelisting mechanism: Through configuration specification, it supports specifying group names and user names as whitelists, thereby realizing controllable scope dialogue with robots
?‍♀️ Administrator mechanism: By configuring designated administrators, some sensitive operations, and some application configurations, the administrator has permission to operate.
㊙️ Sensitive word filtering: By configuring a specified sensitive word, it will be triggered when asking a question, and the answered content will be triggered instead.
? stream mode: Specify DingTalk's stream mode. DingTalk has now fully opened this function, and the project is also started in this mode by default.

Prerequisites for use

If you have an Openai account and create api_key , you can refer to this article for registration related matters. Visit here to apply for a personal key.
Create an application in the DingTalk developer background, add a robot in the application's message push function block, and specify the message receiving mode as stream mode.

Usage tutorial

The first step is to deploy the application

docker deployment

It is recommended that you use docker to run this project quickly.

第一种：基于环境变量运行
# 运行项目
$ docker run -itd --name chatgpt -p 8090:8090 
  -v ./data:/app/data --add-host="host.docker.internal:host-gateway" 
  -e LOG_LEVEL="info" -e APIKEY=换成你的key -e BASE_URL="" 
  -e MODEL="gpt-3.5-turbo" -e SESSION_TIMEOUT=600 
  -e MAX_QUESTION_LENL=2048 -e MAX_ANSWER_LEN=2048 -e MAX_TEXT=4096 
  -e HTTP_PROXY="http://host.docker.internal:15732" 
  -e DEFAULT_MODE="单聊" -e MAX_REQUEST=0 -e PORT=8090 
  -e SERVICE_URL="你当前服务外网可访问的URL" -e CHAT_TYPE="0" 
  -e ALLOW_GROUPS=a,b -e ALLOW_OUTGOING_GROUPS=a,b -e ALLOW_USERS=a,b -e DENY_USERS=a,b -e VIP_USERS=a,b -e ADMIN_USERS=a,b -e APP_SECRETS="xxx,yyy" 
  -e SENSITIVE_WORDS="aa,bb" -e RUN_MODE="http" 
  -e AZURE_ON="false" -e AZURE_API_VERSION="" -e AZURE_RESOURCE_NAME="" 
  -e AZURE_DEPLOYMENT_NAME="" -e AZURE_OPENAI_TOKEN="" 
  -e DINGTALK_CREDENTIALS="your_client_id1:secret1,your_client_id2:secret2" 
  -e HELP="欢迎使用本工具nn你可以查看：[用户指南](https://github.com/eryajf/chatgpt-dingtalk/blob/main/docs/userGuide.md)nn这是一个[开源项目](https://github.com/eryajf/chatgpt-dingtalk/)
  ，觉得不错你可以来波素质三连."  
  --restart=always  registry.cn-hangzhou.aliyuncs.com/eryajf/chatgpt-dingtalk

The configuration file mapped in the run command is referenced below for the configuration file description.

? 注意： If deployed using docker, no adjustments are required for the PORT parameters.
? 注意： The ALLOW_GROUPS,ALLOW_USERS,DENY_USERS,VIP_USERS,ADMIN_USERS parameters are arrays. If multiple parameters need to be specified, they can be divided by English commas. outgoing These parameters are invalid in robot mode.
? 注意： If the server node itself is abroad or has customized BASE_URL , then leave HTTP_PROXY parameter blank.
? 注意： If you use docker to deploy, then the proxy address can be deployed directly in the above method. host.docker.internal will point to the IP of the host host where the container is located. You only need to change the port to your proxy port. See: How Docker containers elegantly access host networks

第二种：基于配置文件挂载运行
# 复制配置文件，根据自己实际情况，调整配置里的内容
$ cp config.example.yml config.yml  # 其中 config.example.yml 从项目的根目录获取

# 运行项目
$ docker run -itd --name chatgpt -p 8090:8090  -v `pwd`/config.yml:/app/config.yml --restart=always  registry.cn-hangzhou.aliyuncs.com/eryajf/chatgpt-dingtalk

The configuration file is referenced to the configuration file description below.

第三种：使用 docker compose 运行
$ wget https://raw.githubusercontent.com/eryajf/chatgpt-dingtalk/main/docker-compose.yml

$ vim docker-compose.yml # 编辑 APIKEY 等信息

$ docker compose up -d

There is also a step to configure Nginx after the deployment is completed. Now the mode is specified as stream mode by default, so Nginx is no longer required.

Binary deployment

If you want to deploy directly through the command line, you can directly download the compressed package in release. Please select the appropriate compressed package according to your system and architecture, and then unzip it directly after downloading.

After downloading, unzip locally and you will see the executable program and configuration file:

$ tar xf chatgpt-dingtalk-v0.0.4-darwin-arm64.tar.gz
$ cd chatgpt-dingtalk-v0.0.4-darwin-arm64
$ cp config.example.yml  config.yml
$ ./chatgpt-dingtalk  # 直接运行

# 如果要守护在后台运行
$ nohup ./chatgpt-dingtalk & > run.log &
$ tail -f run.log

Step 2: Add an application

DingTalk officially launched the stream mode in May 2023, so it is recommended that you use this mode directly here. Other HTTPs are still supported, but they will not conduct in-depth research. Therefore, the following documents are also introduced in the configuration process of the stream mode.

Create steps reference document: internal enterprise application, or configure it according to the following steps.

Create an app.
? Click me to view the sample picture
? 注意： The name of chatgpt may be restricted by DingTalk when creating a robot. Please name it with another name.
In基础信息-->应用信息, you can obtain the robot's AppKey and AppSecret .
Configure the robot.