Open-Source RAG with LLaMa 13B (4 bits for less GPU memory), Faiss, HuggingFace and Langchain or with OpenAI. With gradio UI
Description:
In this Poc we'll create a RAG Open-Source solution with Llama-13b-chat with HuggingFace embedings, Faiss (Vector DB), all orchestrated by LangChain. Or we could parametrize with OpenAI.With gradio UI.
In terms of struture of the solution, we have the main UI in file RAG_QAw_Parametrization.ipynb that import all the parametrization (which model, temperature, chain...) from parametrization.ipynb and the core RAG functions from RAGQA.ipynb. RAGQA.ipynb import also Parametrization.ipynb.
Retrieval Augmented Generation (RAG) is an advanced Natural Language Processing (NLP) technique that combines both retrieval and generation elements to enhance AI language models' capabilities.
We can import your own knowledge database with the solution.
Due to the power of GPU needed i advise you to use colab with RAG_QAw_Parametrization.ipynb(for Llama2). Need to insert your huggingface key
🧮 Virtual Assistant Parametrization UI:
🧮 Virtual Assistant Question Anwsering UI:
Please see the description in .ipynb about this project.
LLama2 with 4096 token on window prompt show a natural improvement in VA Open Source