Add support for Palm, Claude-2, Llama2, CodeLlama (100+LLMs)

[ishaan-jaff]

📝 Description

Description:
This PR adds support for the above mentioned LLMs using LiteLLM https://github.com/BerriAI/litellm/

In addition LiteLLM client allows you to:

• A/B test LLMs in production
• Dynamically control each LLMs prompt, temperature, top_k etc in our UI (no need to re-deploy code)
• Logging to view input/outputs for each LLM

Here's a link to a live demo of litellm client: https://admin.litellm.ai/

Reference Issues: #XX (XX is the issue number you work on)

❕ Linked Issue

👷 Maintainer

@mpskex

[ishaan-jaff]

@mpskex can you take a look at this PR when possible ? happy to add docs/testing too if this initial commit looks good to you

[mpskex]

@mpskex can you take a look at this PR when possible ? happy to add docs/testing too if this initial commit looks good to you

Thanks a LOT for the PR @ishaan-jaff ! Usually we add new models instead of replacing them. So do you mind make a new copy from openai.py to litellm.py?

[ishaan-jaff]

Hi @mpskex curious why make a new file ? LiteLLM allows you to support openai. You can make your openai calls using litellm too

[mpskex]

Sure, LiteLLM is a powerful interface that interacts with many LLMs. But we are trying to keep the compatibility to users and the key philosophy here is have as many options available to users as possible. So we encourage adding instead of replacing, just like langchain did