Error when trying to create a completion with GPT4-preview-vision

In my code, I wanted to add the possibility to use the vision version. When I use gpt4 it works fine, but when I use vision version. I get this error:

openai.BadRequestError: Error code: 400 - {‘error’: {‘message’: ‘1 validation error for Request\nbody → stop\n none is not an allowed value (type=type_error.none.not_allowed)’, ‘type’: ‘invalid_request_error’, ‘param’: None, ‘code’: None}}

Here is a snippet of my code:

            if "vision" in self.config.model_name:
                messages = [
                                "role": "user", 
                                "content": [
                messages = [{"role": "user", "content": prompt}]
            chat_completion =
                            model=self.config["model_name"],  # Choose the engine according to your OpenAI plan
                            max_tokens=n_predict,  # Adjust the desired length of the generated response
                            n=1,  # Specify the number of responses you want
                            stop=None,  # Define a stop sequence if needed
                            temperature=gpt_params["temperature"],  # Adjust the temperature for more or less randomness in the output

OK, it turns out the problem comes from the stop=None. It is not supported on vusion models.

Sorry, can you share what you used as the stop value, can’t seem to hunt it down in the API docs, might be going blind though.