We should implement support for fine-tuned models for chat completion/chatting in general. This would query and select all available models for the user that are classified as fine tunes.

• This will mean we can remove the restriction on model selection on OpenAI LLM selection