maxilloPRO

specialized maxillofacial prosthetic large language model for education and clinical-decision support

Setting up the Environment

Tested on miniconda/25.3.1 and Python 3.10.20

Run

conda env create -f environment.yml

Features

This repository consist of codes used to create a Naive RAG knowledge base alongside with a sample code on creating an LLM-RAG system.

run_build.py

Create the base knowledge base from pdf files.

Current version assumes a 2-column pdf format.

run_append.py

If you wish to extend the knowledge base with new sources or manually curated ones, you can use this code to do so.

run_query.py

A sample code on how to extract relevant context based on cosine similarity between User query and all documents.

run_rag_llm.py

A sample code on how to couple the RAG's output as the input for the LLM.