app.py

import os
import faiss
import streamlit as st
from PyPDF2 import PdfReader
from sentence_transformers import SentenceTransformer
from groq import Groq
from dotenv import load_dotenv



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

# Load Sentence Transformer model
model = SentenceTransformer("all-MiniLM-L6-v2")

# Initialize FAISS
dimension = 384  # Embedding size for the Sentence Transformer model
index = faiss.IndexFlatL2(dimension)

# Function to process PDF and create embeddings
def process_pdf(pdf_file):
    pdf_reader = PdfReader(pdf_file)
    text = ""
    for page in pdf_reader.pages:
        text += page.extract_text()
    chunks = [text[i:i + 500] for i in range(0, len(text), 500)]  # Chunk into 500-char blocks
    embeddings = model.encode(chunks)
    index.add(embeddings)
    return chunks, embeddings

# Function to query FAISS and generate a response
def query_model(query):
    query_vector = model.encode([query])
    _, indices = index.search(query_vector, k=3)  # Top 3 similar chunks
    response_chunks = [stored_chunks[idx] for idx in indices[0]]
    context = " ".join(response_chunks)

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

# Streamlit app
st.title("RAG-based PDF Question Answering")
st.write("Upload a PDF and ask questions based on its content.")

uploaded_file = st.file_uploader("Upload your PDF", type=["pdf"])
if uploaded_file:
    stored_chunks, _ = process_pdf(uploaded_file)
    st.success("PDF processed and embeddings created.")

    query = st.text_input("Ask a question:")
    if query:
        answer = query_model(query)
        st.write("### Answer:")
        st.write(answer)