[FEAT]: Huggingface Serverless Inference API

[CRTTV]

What would you like to see?

Hello! I was looking for a RAG program that would allow me to use my school textbooks. I was using RAGflow and Open WebUI but found I couldn't use Huggingface on RAGflow and RAG is lacking on Open WebUI.

I saw a post mentioning Anything LLM and downloaded the program and it seems really good! I'm having trouble with adding huggingface to it and using the Serverless Inference API.

The page for huggingface on Anything LLM is was not helpful to me as there no usable information there:

https://docs.useanything.com/setup/llm-configuration/cloud/hugging-face#connecting-to-hugging-face

Is there a way to add the serverless inference API models and use those on Anything LLM? I use the Huggingface model URL (https://api-inference.huggingface.co/models/Qwen/Qwen2.5-72B-Instruct/v1/chat/completions in my case) on the Inference Endpoint box but it says I need a .cloud URL instead.

Thank you for any help!