Getting AttributeError when using openAI python library

题意:"使用 OpenAI Python 库时出现 AttributeError"

问题背景:

I'm building a new AI chatbot utilizing the openai library and I have a gradio UI set up in one file (app.py) and a predict() function in another (trainedBot.py) Every time I send a request through the gradio UI I get this error:

"我正在使用 OpenAI 库构建一个新的 AI 聊天机器人,并在一个文件 (app.py) 中设置了 Gradio 界面,在另一个文件 (trainedBot.py) 中编写了 predict() 函数。每次我通过 Gradio 界面发送请求时都会出现这个错误:"

python 复制代码
File "/home/user/app/trainedBot.py", line 48, in predict
    return response.choices.message.content
AttributeError: 'list' object has no attribute 'message'

I tried putting both the message and the openAI response in a variable and nothing happened. Still got the same error. The response looked like this:

"我尝试将消息和 OpenAI 的响应都放在一个变量中,但没有任何变化。仍然出现相同的错误。响应看起来是这样的:"

python 复制代码
{
  "id": "chatcmpl-8CHKgpOewruWDC2Et1R6ZFtdPmSQR",
  "object": "chat.completion",
  "created": 1697937254,
  "model": "gpt-4-0613",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "HIIIIIII"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 308,
    "completion_tokens": 4,
    "total_tokens": 312
  }
}

问题解决:

python 复制代码
response = {
  "id": "chatcmpl-8CHKgpOewruWDC2Et1R6ZFtdPmSQR",
  "object": "chat.completion",
  "created": 1697937254,
  "model": "gpt-4-0613",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "HIIIIIII"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 308,
    "completion_tokens": 4,
    "total_tokens": 312
  }
}

Your choices is a list so access it you need to put [0]

"你的 `choices` 是一个列表,所以要访问它,你需要使用 `[0]`。"

python 复制代码
print(response['choices'][0]['message']['content'])

'HIIIIIII'

Or could be: 或者也可以用以下的方式

python 复制代码
response.get('choices')[0].get('message').get('content')
相关推荐
安逸sgr7 分钟前
循环神经网络 RNN、LSTM、GRU 是什么?为什么能处理序列?
人工智能·ai·大模型·agent·智能体
猿小猴子18 分钟前
主流 AGENT 实战教程「OpenCodex」与「ChatGPT Work」与「Codex Record & Replay」介绍
人工智能·ai·chatgpt·codex·minimax·deepseek·opencodex
阿标的博客31 分钟前
Python实训案例(二):输出植物证书
python
鸿芯微控科技35 分钟前
压电点胶阀出胶量不稳定怎么排查?驱动波形、背压、点胶重量与Python分析
开发语言·python·压电点胶阀·精密点胶·点胶重量·流体控制
一位正在转型AI全栈的前端工程师37 分钟前
从 0 到 1 搭建生产级 RAG 系统:切片、召回与重排序调优实录
python·前端工程化
糖炒栗子03261 小时前
learn-claude-code 简要记录
笔记·python
TechEdu2026061 小时前
[人工智能]RLlib:基于Ray的可扩展强化学习框架
人工智能·ai
测试19981 小时前
Python接口测试之requests库安装和导入
自动化测试·软件测试·python·测试工具·职场和发展·测试用例·接口测试
%471 小时前
DAY 34
python
leisoo80971 小时前
涨停板次日表现因子怎么挖掘本地化Python全流程实战
大数据·人工智能·python