Have you ever wondered, if there's a way to summarise your docs obtain answers from a knowledge base that may not be accessible to chat assistants? We've addressed that challenge for you! With "Chat-with-your-docs", you can simply upload your document or provide a website link. Our assistant will be able to answers based on the provided documents in no time.
Don't forget to star our repository if liked it :) .
Before you begin, this repository uses LLM models and embedding models from hugging face through inference api. You need to have huggingface token to move forward. It is free of cost and can be created here Link .
Clone repository:
git clone https://github.com/kartheekyakkala/Chat-with-your-docs.gitInstall poetry:
pip install poetryInstall dependencies:
poetry installThats'it you can run the main file to start chat.
poetry run python chat-with-your-docs/main.pyThis repo uses LangChain as framework to integrate with open source LLM and Retreival Augmented Generation (RAG) Technique to add context from knowledge base as prompt to the LLM.
If you're interested in learning more about RAG, you can explore my blog post by clicking here .