Support to local API model

[fcencio]

Hi,

I've deployed a llamav2 locally, and it's exposed at localhost:5000/api/v1/generate
I cannot understand how to make this model availables on openpalyground. Is it possible to do this?

Thank you

[krrishdholakia]

Why not have it replace the openai.api_base?

You might need to change your query endpoint to being /chat/completions . Here's an openai proxy server template that might help - https://github.com/BerriAI/litellm/tree/main/proxy-server