github.com/yoak3n/EdgeGPT-http

EdgeGPT-http

a http server based on go

just a practice work for self-entertainment.

Download

git clone https://github.com/Yoak3n/EdgeGPT-http.git

or

Download ZIP

Usage

Basic(manually)

Find the configuration file config.example.yaml and rename it as config.yaml, then modify the configuration inside.

• Start the service by using the following command:

go run .\main.go

• Client post to url/chat with a json data like the following:

{
  "name": "test",
  "style": "bing-c",
  "question": "hello bing"
}

One namemeans one conversation
Then the client will get a json response:

{
  "status":  "success",
  "style":   "bing-c",
  "message": "bing's answer",
  "count": {
    "currentNum":1,
    "maxNum":20
  }
}

Docker

Must installed Docker

• Pull the image by:

docker pull yoaken/edgegpt-http:latest 

2.Run the image:

docker run -v 'your absolute path of config.yml:/app/config.yml' -v'your absolute path of cookies.json:/app/cookies.json' -p "8080:8080" yoaken/edgegpt-http:latest 

docker-compose(recommended)

Must installed Docker and docker-compose

• Copy the file docker-compose.yml to the dir path as you like
• Create your config.yml and cookies.json in the path the same to the above
• Change the work dir where these files exist and run with a single command:

docker-compose up -d  

Problem

• multiple sessions
• commands for conversation like resetor more
• source auto release
• high concurrency(perhaps gin already supported)

Reference

At the very beginning I use the Python lib EdgeGPT in my bing-qqbot,it's too complicated to start.So I planed to convert it to golang at some point,however thank billikeu that his repository Go-EdgeGPT really helped me(btw my plan is also named go-EdgeGPT)XD
For the convenience of development, it is temporarily included as part of this project, and may be submitted as a PR in the future

And the idea came from the http service of chatgpt-mirai-qq-bot