I see at https://platform.openai.com/docs/assistants/tools/code-interpreter that I can use a fileId in the assistants API. However, I am shy of doing this in a production environment as assistant API is still in beta.
I would like to use the API for chatGPT (preferably4 , or worst case3.5 turbo) to reply to me based on a fileId that I would have uploaded earlier.
Is this possible at all?
What else should I include in my CURL call to OpenAi API when i call https://api.openai.com/v1/chat/completions (i mean along with model, stream, messages, etc.)
Thanks a ton in advance!