src/app.py

# Databricks notebook source
import streamlit as st
import os
import yaml
from dotenv import load_dotenv
from src.generator import answer_with_rag
from ragatouille import RAGPretrainedModel
from src.data_preparation import split_documents
from transformers import pipeline
from langchain_community.document_loaders import PyPDFLoader
from langchain.embeddings import HuggingFaceEmbeddings
from src.retriever import init_vectorDB_from_doc, retriever
from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig
from langchain_community.vectorstores import FAISS
import faiss
def load_config():
    with open("./src/config.yml","r") as file_object:
        try:
            cfg=yaml.safe_load(file_object)
            
        except yaml.YAMLError as exc:
            logger.error(str(exc))
            raise
        else:
            return cfg

cfg= load_config()
load_dotenv("./src/.env")

EMBEDDING_MODEL_NAME=cfg['EMBEDDING_MODEL_NAME']
DATA_FILE_PATH=cfg['DATA_FILE_PATH']
READER_MODEL_NAME=cfg['READER_MODEL_NAME']
RERANKER_MODEL_NAME=cfg['RERANKER_MODEL_NAME']
VECTORDB_PATH=cfg['VECTORDB_PATH']
if __name__ == "__main__":
    st.title("RAG App to query le College de Pédiatrie")

    user_query = st.text_input("Entrez votre question:")


    # Initialize the retriever and LLM
    
    loader = PyPDFLoader(DATA_FILE_PATH)
    #loader = PyPDFDirectoryLoader(DATA_FILE_PATH)
    raw_document_base = loader.load()
        MARKDOWN_SEPARATORS = [
        "\n#{1,6} ",
        "```\n",
        "\n\\*\\*\\*+\n",
        "\n---+\n",
        "\n___+\n",
        "\n\n",
        "\n",
        " ",
        "",]
    docs_processed = split_documents(
        512,  # We choose a chunk size adapted to our model
        raw_document_base,
        tokenizer_name=EMBEDDING_MODEL_NAME,
        separator=MARKDOWN_SEPARATORS
    )
    embedding_model=init_embedding_model(EMBEDDING_MODEL_NAME)
    
    if os.path.exists(VECTORDB_PATH):
        new_vector_store = FAISS.load_local(
            VECTORDB_PATH, embedding_model, 
            allow_dangerous_deserialization=True)
    else:
        KNOWLEDGE_VECTOR_DATABASE=init_vectorDB_from_doc(docs_processed, embedding_model)
        KNOWLEDGE_VECTOR_DATABASE.save_local(VECTORDB_PATH)


    if st.button("Get Answer"):
    # Get the answer and relevant documents
        bnb_config = BitsAndBytesConfig(
            load_in_4bit=True,
            bnb_4bit_use_double_quant=True,
            bnb_4bit_quant_type="nf4",
            bnb_4bit_compute_dtype=torch.bfloat16,
        )
        model = AutoModelForCausalLM.from_pretrained(READER_MODEL_NAME, quantization_config=bnb_config)
        tokenizer = AutoTokenizer.from_pretrained(READER_MODEL_NAME)

        READER_LLM = pipeline(
            model=model,
            tokenizer=tokenizer,
            task="text-generation",
            do_sample=True,
            temperature=0.2,
            repetition_penalty=1.1,
            return_full_text=False,
            max_new_tokens=500,
        )
        RERANKER = RAGPretrainedModel.from_pretrained(RERANKER_MODEL_NAME)
        num_doc_before_rerank=15
        num_final_releveant_docs=5
        answer, relevant_docs = answer_with_rag(query=user_query, READER_MODEL_NAME=READER_MODEL_NAME,embedding_model=embedding_model,vectorDB=KNOWLEDGE_VECTOR_DATABASE,reranker=RERANKER, llm=READER_LLM,num_doc_before_rerank=num_doc_before_rerank,num_final_relevant_docs=num_final_releveant_docs,rerank=True)
    #print(answer) 

    
    # Display the answer
        st.write("### Answer:")
        st.write(answer)
    
    # Display the relevant documents
        st.write("### Relevant Documents:")
        for i, doc in enumerate(relevant_docs):
            st.write(f"Document {i}:\n{doc.text}")