Consider adding support for gpt-3.5-turbo-instruct model

[smuotoe]

Using the instruct model raises an error.

ai = AIChat(console=False, model="gpt-3.5-turbo-instruct")

KeyError: "No AI generation: {'error': {'message': 'This is not a chat model and thus not supported in the v1/chat/completions endpoint. Did you mean to use v1/completions?', 'type': 'invalid_request_error', 'param': 'model', 'code': None}}"

[minimaxir]

So I won't add support for this due to the reason gpt-3.5-turbo-instruct exists: it was added to accomidate the soon-to-be-depreciated GPT-3 models. As a result, it has a different API that's othogonal to the current API simpleaichat uses, so it will increase scope for not much gain.

OpenAI doesn't intend for its use in day-to-day operations compared to its ChatGPT APIs.

[minimaxir]

Actually it may be better to add an explicit assert dictating this.