OpenAi not returning result and exited with code=0

news2024/11/17 7:17:15

题意：OpenAi 没有返回结果，程序以代码 0 退出

问题背景：

I'm trying to use OpenAI, but I can't get a result. I'm accessing the API via Visual Studio Code. I have downloaded these extensions for Visual Studio Code: Code Runner and Python. I've also installed OpenAI via CMD: pip install openai.

你正在尝试使用 OpenAI，但是无法获取结果。你通过 Visual Studio Code 访问 API，并且已经为 Visual Studio Code 下载了这些扩展：Code Runner 和 Python。此外，你还通过 CMD 安装了 OpenAI 的 Python SDK，即使用 pip install openai 命令。

Here's my code: 这是我的代码：

import os
import openai
openai.api_key = os.getenv("sk-xxxxxxxxxxxxxxxxxxxx")
x=openai.Completion.create(
  engine="text-davinci-002",
  prompt="Say this is a test",
  max_tokens=5
)
print(x)

Referenced from the official documentation:

引用自官方文档：

https://beta.openai.com/docs/api-reference/completions/create?lang=python

But when I run that code, the output tab is not outputting anything:

但是当我运行那段代码时，输出标签没有输出任何东西：

Anyone know where I possibly went wrong?

有人知道我可能哪里出错了吗？

问题解决：

When I run your code in console/terminal/bash (on Linux) without VSCode then I get some useful error message. So maybe first you should test it on CMD to see if you get error message with explanation.

当我在 Linux 上的控制台/终端/bash 中运行你的代码（不使用 VSCode），我得到了一些有用的错误消息。所以，也许你应该首先在 CMD 中测试一下，看看是否会出现带有说明的错误消息。

But main problem is that you use API_KEY in wrong way

但是主要问题是你错误地使用了 API_KEY

You should use it directly in code (without os.getenv())

你应该直接在代码中使用它（不使用 os.getenv()）

openai.api_key = "sk-5kyIzSG6wxeCDdf2T3BlbdfJxgdfeet9JWm8cQumrG"

Or in system you should set environment's variable

或者你应该在系统中设置环境变量

OPEN_API_KEY=sk-5kyIzSG6wxeCDdf2T3BlbdfJxgdfeet9JWm8cQumrG

and use exactly OPEN_API_KEY

并使用确切的 OPEN_API_KEY

openai.api_key = os.getenv("OPEN_API_KEY")

(and this way you can share code without sharing API_KEY)

这样你就可以在不共享 API_KEY 的情况下共享代码

Your API_KEY is too short but I tested it with my API_KEY and it works for me.

你的 API_KEY 太短了，但我用我自己的 API_KEY 测试过，它对我来说是有效的。

import openai

openai.api_key = "sk-...my_api_key..."

x = openai.Completion.create(
  engine="text-davinci-002",
  prompt="Say this is a test",
  max_tokens=5
)

print(x)

Result: 结果：

{
  "choices": [
    {
      "finish_reason": "length",
      "index": 0,
      "logprobs": null,
      "text": "\n\nThis is a"
    }
  ],
  "created": 1652054180,
  "id": "cmpl-55l36Li5BTrRZWPU38MdQai8yVGEA",
  "model": "text-davinci:002",
  "object": "text_completion"
}