Setting up the ChatGPT API

Hi all,

I’m trying to setup the chatGPT API, however I’m getting the following error:
InvalidRequestError : Invalid URL (POST /v1/chat/completions/engines/gpt-3.5-turbo/chat/completions

I’m using the following code:

messages = [
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Who won the world series in 2020?"},
        {"role": "assistant", "content": "The Los Angeles Dodgers won the World Series in 2020."},
        {"role": "user", "content": "Where was it played?"}
        ]

response = openai.ChatCompletion.create(
            engine="gpt-3.5-turbo",
            messages=messages)

Any ideas what I could be doing wrong

UPDATE
engine was supposed to be model

1 Like

/v1/chat/completions/engines/gpt-3.5-turbo/chat/completions

Is completely wrong. Maybe try updating the openai module?

1 Like

I’m on openai version 0.27.0, I believe this is the latest one right? I’ll try reinstalling it

UPDATE:
Tried reinstalling it with no avail… Anyone has suggestions?

You need to use …

https://api.openai.com/v1/chat/completions

… and send a model parameter.

I’m thinking the latest OpenAI module likely has it… I think 0.27 is old?

Thanks for the reply, but how do I use the link you provided? Should I change openai.api_base? I’m fairly new to this. To add, the error now is: InvalidRequestError : Invalid URL (POST /v1/engines/gpt-3.5-turbo/chat/completions), which is slightly different from before

PS. I checked on the openai github and 0.27.0 is the latest version

1 Like

How big is your file? Could you show all of your code, including the imports?

1 Like

As per PaulBellow’s suggestion, try changing the key engine to model.

messages = [
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Who won the world series in 2020?"},
        {"role": "assistant", "content": "The Los Angeles Dodgers won the World Series in 2020."},
        {"role": "user", "content": "Where was it played?"}
        ]

response = openai.ChatCompletion.create(
            model="gpt-3.5-turbo",
            messages=messages)

Slightly confusing interface change, if you ask me.

2 Likes

Hi,

Changing the engine to model worked. Thank you all!

1 Like

I got the same error when request the API “/v1/chat/completions”,have you solve the error?

Hi,

This code resolved the issue for me.

import requests
import openai
messages = [
{“role”: “user”, “content”: “What is you name?”}
]
url = ‘https://api.openai.com/v1/chat/completions
headers = {“Authorization”: f"Bearer {api_key}"}
data = {‘model’: ‘gpt-3.5-turbo’, ‘messages’: messages}
requests.post(url, headers=headers, json=data).json()