Is the library compatible with LLMs aside from OpenAI?

[muhammedyusuf678]

I have been trying to run validation over the Dolly-3b model from the HuggingFaceHub via the Langchain API. Since this library integrates with langchain, I was assuming it would work. However, I only always get None outputs. I am also curious what the engine parameter of the guard function represents if the LLM is not from openAI.

My LLM model code:
`from langchain.llms import HuggingFaceHub
from langchain.chains import LLMChain

repo_id = "databricks/dolly-v2-3b"
llm = HuggingFaceHub(repo_id=repo_id)

raw_llm_output, validated_output = guard(
llm_api=llm,
# engine="text-davinci-003"
)`

[irgolic]

Hi, sorry for the late reply, please see here for a guide on how to use custom LLMs: https://docs.guardrailsai.com/llm_api_wrappers/#using-a-custom-llm-api

Let me know if you have any further questions.

[irgolic]

If you run into problems with Dolly, feel free to send over an example of your data and rail spec :)