app.py

import streamlit as st
import os
from langchain_groq import ChatGroq
from langchain_huggingface import HuggingFaceEmbeddings
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.chains.combine_documents import create_stuff_documents_chain
from langchain_core.prompts import ChatPromptTemplate
from langchain.chains import create_retrieval_chain
from langchain_community.vectorstores import FAISS
from langchain_community.document_loaders import PyPDFLoader
from dotenv import load_dotenv
import tempfile


load_dotenv()


groq_api_key = os.getenv('GROQ_API_KEY')


st.markdown("<h2 style='text-align: center;'>PDF Insights: Interactive Q&A Chatbot with Groq API</h2>", unsafe_allow_html=True)


llm = ChatGroq(groq_api_key=groq_api_key, model_name="Llama3-8b-8192")


prompt = ChatPromptTemplate.from_template(
    """
    Answer the questions based on the provided context only.
    Please provide the most accurate response based on the question.
    If the answer is not in the document, just say "Please Contact the Business Directly". Dont say wrong answer.
    <context>
    {context}
    <context>
    Questions: {input}
    """
)

def create_vector_db_out_of_the_uploaded_pdf_file(pdf_file):


    if "vector_store" not in st.session_state:

        with tempfile.NamedTemporaryFile(delete=False) as temp_file:

            temp_file.write(pdf_file.read())

            pdf_file_path = temp_file.name

        st.session_state.embeddings = HuggingFaceEmbeddings(model_name='BAAI/bge-small-en-v1.5', model_kwargs={'device': 'cpu'}, encode_kwargs={'normalize_embeddings': True})
        
        st.session_state.loader = PyPDFLoader(pdf_file_path)

        st.session_state.text_document_from_pdf = st.session_state.loader.load()

        st.session_state.text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
        
        st.session_state.final_document_chunks = st.session_state.text_splitter.split_documents(st.session_state.text_document_from_pdf)

        st.session_state.vector_store = FAISS.from_documents(st.session_state.final_document_chunks, st.session_state.embeddings)


pdf_input_from_user = st.file_uploader("Upload the PDF file", type=['pdf'])


if pdf_input_from_user is not None:

    if st.button("Create the Vector DB from the uploaded PDF file"):
        
        if pdf_input_from_user is not None:
            
            create_vector_db_out_of_the_uploaded_pdf_file(pdf_input_from_user)
            
            st.success("Vector Store DB for this PDF file Is Ready")
        
        else:
            
            st.write("Please upload a PDF file first")



if "vector_store" in st.session_state:

    user_prompt = st.text_input("Enter Your Question related to the uploaded PDF")

    if st.button('Submit Prompt'):

        if user_prompt:
            
            if "vector_store" in st.session_state:

                document_chain = create_stuff_documents_chain(llm, prompt)

                retriever = st.session_state.vector_store.as_retriever()

                retrieval_chain = create_retrieval_chain(retriever, document_chain)

                response = retrieval_chain.invoke({'input': user_prompt})

                st.write(response['answer'])

            else:

                st.write("Please embed the document first by uploading a PDF file.")

        else:

            st.error('Please write your prompt')