added system prompt for openai

[Leo-LiHao]

What does this PR do?

Add system prompt for OpenAI Representation model, see Issue #2146

Before submitting

• This PR fixes a typo or improves the docs (if yes, ignore all other checks!).
• Did you read the contributor guideline?
• Was this discussed/approved via a Github issue? Please add a link to it if that's the case.
• Did you make sure to update the documentation with your changes (if applicable)?
• Did you write any new necessary tests?

[MaartenGr]

Thanks for updating these two models! I left a small comment.

Other than that, I believe we will need to look at the following:

• HuggingFace
• LangChain
• Llama-cpp-python

With HuggingFace, we might need to change the formatting such that we use messages instead:

messages = [
    {
        "role": "system",
        "content": MY_SYSTEM_PROMPT,
    },
    {"role": "user", "content": prompt},
 ]

For LangChain, I'm okay with skipping this for now since it uses a more complex structure and the API here might need to be updated (see https://python.langchain.com/v0.1/docs/modules/model_io/chat/quick_start/#messages-in---message-out).

For llama-cpp-python, it seems straightforward and we should use .create_chat_completion instead of directly calling the model.

[MaartenGr]

Why did you make this change? In my experience, by providing the model with a pre-fix, there is no need for mentioning that it should provide the topic name without any explanation.

[Leo-LiHao]

Based on my experiments, the pre-fix prompt works well for completion models but not with (small) chat models. You can try different prompts locally to see the difference and feel free to change/iterate the prompt.

[Leo-LiHao]

@MaartenGr I updated the code for llama-cpp