README.md

metadata

license: mit
title: Medical Document Assistant APP with LLM RAG framework
sdk: docker
emoji: 📚
colorFrom: blue
colorTo: red
pinned: false
short_description: Search medical terms among uploaded document
app_port: 8080

Necessary resources

Model must be downloaded to local ai_workshop folder:

Llama 2 Model (Quantized one by the Bloke): https://huggingface.co/TheBloke/Llama-2-7B-Chat-GGML/blob/main/llama-2-7b-chat.ggmlv3.q8_0.bin

License and other reference

The code in all scripts subjects to a licence of 96harsh52/LLaMa_2_chatbot (https://github.com/96harsh52/LLaMa_2_chatbot) Youtube instruction (https://www.youtube.com/watch?v=kXuHxI5ZcG0&list=PLrLEqwuz-mRIdQrfeCjeCyFZ-Pl6ffPIN&index=18)

Llama 2 HF Model (Original One): https://huggingface.co/meta-llama Chainlit docs: https://github.com/Chainlit/chainlit

Create virtual Environment

1. Create Virtual env:

   cd ai_workshop python -m venv langchain

2. Activate virtual evn:

   langchain\Scripts\activate

*NOTE: if you see the read warning in cmd terminal said "running scripts is disabled on this system" , use Powershell to setup API server: 1. open Powershell > Set-ExecutionPolicy Unrestricted -Scope Process 2. activate virtual env as previous steps

3. install requirements.txt

   python -m ensurepip --upgrade python -m pip install --upgrade setuptools python -m pip install -r requirements.txt

Create local vectors storage database

After activate virtual environment, run python .\ingest.py

Setup Medical chatbot server with chainlit

After set up the database folder of "vectorstore/db_faiss", run chainlit run .\model.py > logs.txt