app.py

import os
from groq import Groq
from langchain_community.embeddings import HuggingFaceEmbeddings
from langchain_community.vectorstores import FAISS
from langchain.text_splitter import RecursiveCharacterTextSplitter
from PyPDF2 import PdfReader
import streamlit as st
from tempfile import NamedTemporaryFile

# Initialize Groq client
client = Groq(api_key="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")

# Function to extract text from a PDF
def extract_text_from_pdf(pdf_file_path):
    pdf_reader = PdfReader(pdf_file_path)
    text = ""
    for page in pdf_reader.pages:
        text += page.extract_text()
    return text

# Function to split text into chunks
def chunk_text(text, chunk_size=500, chunk_overlap=50):
    text_splitter = RecursiveCharacterTextSplitter(
        chunk_size=chunk_size, chunk_overlap=chunk_overlap
    )
    return text_splitter.split_text(text)

# Function to create embeddings and store them in FAISS
def create_embeddings_and_store(chunks):
    embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
    vector_db = FAISS.from_texts(chunks, embedding=embeddings)
    return vector_db

# Function to query the vector database and interact with Groq
def query_vector_db(query, vector_db):
    # Retrieve relevant documents
    docs = vector_db.similarity_search(query, k=3)
    context = "\n".join([doc.page_content for doc in docs])

    # Interact with Groq API
    chat_completion = client.chat.completions.create(
        messages=[
            {"role": "system", "content": f"Use the following context:\n{context}"},
            {"role": "user", "content": query},
        ],
        model="llama3-8b-8192",
    )
    return chat_completion.choices[0].message.content

# Streamlit app
st.set_page_config(
    page_title="Auto Buddy: RAG Application",
    page_icon="💻",
    layout="wide",
    initial_sidebar_state="expanded",
)

st.title("📚 Auto Buddy: Your RAG-Powered Assistant")
st.markdown(
    """
    Welcome to **Auto Buddy**, your AI-powered assistant that leverages **Retrieval-Augmented Generation (RAG)** for powerful insights. 
    Upload your PDF documents, ask questions, and receive precise answers effortlessly.
    """
)

# Sidebar Instructions
st.sidebar.header("Instructions")
st.sidebar.write(
    "1. Upload a PDF document.\n"
    "2. Wait for the text extraction and chunking process.\n"
    "3. Enter your query to receive AI-driven answers."
)

# Upload PDF
uploaded_file = st.file_uploader("Upload a PDF Document", type=["pdf"])

if uploaded_file:
    with NamedTemporaryFile(delete=False, suffix=".pdf") as temp_file:
        temp_file.write(uploaded_file.read())
        pdf_path = temp_file.name

    # Extract text
    st.subheader("Step 1: Text Extraction")
    text = extract_text_from_pdf(pdf_path)
    st.success("PDF Text Extracted Successfully!")

    # Chunk text
    st.subheader("Step 2: Text Chunking")
    chunks = chunk_text(text)
    st.success("Text Chunked Successfully!")

    # Generate embeddings and store in FAISS
    st.subheader("Step 3: Embeddings and Storage")
    vector_db = create_embeddings_and_store(chunks)
    st.success("Embeddings Generated and Stored Successfully!")

    # User query input
    st.subheader("Step 4: Ask Your Question")
    user_query = st.text_input("The issue with my car is:")
    if user_query:
        response = query_vector_db(user_query, vector_db)
        st.subheader("Response from LLM")
        st.write(response)