app.py

import streamlit as st
import PyPDF2
import faiss
import numpy as np
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.vectorstores import FAISS
from langchain.llms import OpenAI
from langchain.chains import RetrievalQA

def extract_text_from_pdf(pdf_file):
    reader = PyPDF2.PdfReader(pdf_file)
    text = ''
    for page in reader.pages:
        text += page.extract_text()
    return text

def create_embeddings(text):
    embedding_model = HuggingFaceEmbeddings(model_name="all-MiniLM-L6-v2")
    embeddings = embedding_model.embed_documents([text])
    return embeddings

def create_faiss_index(embeddings):
    dim = len(embeddings[0])
    index = faiss.IndexFlatL2(dim)
    embeddings_np = np.array(embeddings).astype('float32')
    index.add(embeddings_np)
    return index

def create_retrieval_chain(index):
    llm = OpenAI(model="gpt-3.5-turbo")
    vector_store = FAISS(index)
    retrieval_qa = RetrievalQA(combine_docs_chain=llm, vectorstore=vector_store)
    return retrieval_qa

def retrieve_and_generate(query, retrieval_qa):
    response = retrieval_qa.run(query)
    return response

def main():
    st.title("RAG Application with FAISS & PDF")
    
    pdf_file = st.file_uploader("Upload your PDF document", type="pdf")
    
    if pdf_file is not None:
        text = extract_text_from_pdf(pdf_file)
        st.subheader("Extracted Text from PDF")
        st.write(text[:1000])  # Display first 1000 characters for preview

        embeddings = create_embeddings(text)
        index = create_faiss_index(embeddings)
        retrieval_qa = create_retrieval_chain(index)
        
        query = st.text_input("Enter your query:")
        
        if query:
            response = retrieve_and_generate(query, retrieval_qa)
            st.subheader("Answer from RAG Model:")
            st.write(response)

if __name__ == "__main__":
    main()