azure - Azure OpenAI API 的 Google Sheet Apps 脚本自定义函数

Question

我在 Google App 脚本中创建了一个自定义函数来调用 OpenAI API。它运行正常，但我目前正在过渡到 Azure OpenAI API 服务，以提高性能和稳定性。但是，我在尝试根据 Azure OpenAI 说明重新编码 Apps 脚本时遇到了问题。当前脚本版本如下:

const api_key = "azure openai api key1";  
const max_tokens = 500;  
  
function gpt(userprompt, userrole = "user", temperature = 1) {  
  
  const gptprompt = [{role:userrole,content:userprompt},  
  ];  
  
  const url = "azure openai endpoint";  
  const payload = {  
    engine:"gpt-35-turbo",  
    messages: gptprompt,
    temperature: temperature,  
    max_tokens: max_tokens,
  };  
  
  // Use 'method' property with value 'POST' instead of default 'GET' method  
  const options = {  
    method: "POST",  
    contentType: "application/json",  
    headers: {  Authorization: "Bearer " + api_key },  
    payload: JSON.stringify(payload),  
    muteHttpExceptions: true  
  };  
  
  try {  
    const response = UrlFetchApp.fetch(url, options);  
    console.log(response.getContentText());  
    const results = JSON.parse(response.getContentText());  
    console.log(results);  
    const text = results?.choices?.[0]?.message?.text?.trim() ?? "No text generated"; // Update this line to use 'choices' and 'message'  
    console.log(text);  
    return text;  
  } catch (error) {  
    // Log error message and return a custom error message  
    console.error("Error: ", error);  
    return "Error: Unable to fetch data.";  
  }  
}  

但根据日志返回错误:

6:52:23AM 通知执行开始

6:52:23AM Info { "statusCode": 401, "message": "未经授权。访问 token 丢失、无效、受众不正确 ( https://cognitiveservices.azure.com ) 或已过期。"}

6:52:23AM 信息 { statusCode: 401, message: '未经授权。访问 token 丢失、无效、受众不正确 ( https://cognitiveservices.azure.com ) 或已过期。 }

6:52:23AM 信息 未生成任何文本 6:52:24AM 通知执行已完成

我已经检查了 API key 和端点在 python 中是否工作我错过了应用程序脚本中的某些部分吗？

Answer 1

根据 Azure 指南，API key 的 header 已更改。

所以，尝试“api-key: YOUR_API_KEY”

而不是“授权:持有者 YOUR_API_KEY”

#Note: The openai-python library support for Azure OpenAI is in preview.

import os
import openai
openai.api_type = "azure"
openai.api_base = "YOUR_API_ENDPOINT"
openai.api_version = "2023-03-15-preview"
openai.api_key = os.getenv("OPENAI_API_KEY")

response = openai.ChatCompletion.create(
  engine="YOUR_MODEL_NAME_DEPLOYED",
  messages = [{"role":"system","content":"You are an AI assistant that helps people find information."},{"role":"user","content":"Hi"},{"role":"assistant","content":"Hi! What can I do for you?"}],
  temperature=0.7,
  max_tokens=800,
  top_p=0.95,
  frequency_penalty=0,
  presence_penalty=0,
  stop=None)

`