调用智谱清言的图像识别接口

github地址：https://github.com/LLM-Red-Team/glm-free-api

1.docker 来取镜像并部署

  拉取镜像

docker pull vinlic/glm-free-api:latest

 docker部署

docker run -it -d --init --name glm-free-api -p 8000:8000 -e TZ=Asia/Shanghai vinlic/glm-free-api:latest

查看服务实时日志

docker logs -f glm-free-api

重启服务

docker restart glm-free-api

停止服务

docker stop glm-free-api

Docker-compose部署

version: '3'

services:
  glm-free-api:
    container_name: glm-free-api
    image: vinlic/glm-free-api:latest
    restart: always
    ports:
      - "8000:8000"
    environment:
      - TZ=Asia/Shanghai

 

2.接口调用

  demo 代码（图像识别 ）

import requests
import json

url = "http://10.101.92.218:8000/v1/chat/completions"

payload = json.dumps({
    "model": "65c046a531d3fcb034918abe",
    "messages": [
        {
            "role": "user",
            "content": [
                {
                    "type": "image_url",
                    "image_url": {
                        "url": "http://10.101.92.218:9000/456.jpg"
                    }
                },
                {
                    "type": "text",
                    "text": "整理图片内容"
                }
            ]
        }
    ],
    "stream": False
})
headers = {
    'Authorization': 'xxxx',
    'Content-Type': 'application/json'
}

response = requests.request("POST", url, headers=headers, data=payload)

data = response.json()

print(data.get("choices")[0].get("message").get("content"))

  注意点：1.Authorization的值自己去浏览器找   

                2.图片的url--互联网可读取的图片url，可以图片传到某台服务器，然后开个 python -m http.server 9000 让接口可访问