app.py

import streamlit as st
import fitz  # PyMuPDF
from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline

# Load model directly
model_name = "openai-community/gpt2"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)

# Function to extract text from PDF
def extract_text_from_pdf(pdf_file):
    pdf_document = fitz.open(stream=pdf_file.read(), filetype="pdf")
    text = ""
    for page_num in range(pdf_document.page_count):
        page = pdf_document.load_page(page_num)
        text += page.get_text()
    return text

# Function to generate questions using GPT-2
def generate_questions(text, num_questions=5):
    if not text.strip():
        return ["No text extracted from the PDF. Unable to generate questions."]
    
    # Create the text generation pipeline
    generator = pipeline("text-generation", model=model, tokenizer=tokenizer)

    questions = []
    for _ in range(num_questions):
        # Generate a single question at a time
        prompt = f"Generate a question based on the following text:\n{text}\n\nQuestion:"
        generated = generator(prompt, max_length=500, num_return_sequences=1)
        question = generated[0]["generated_text"].split("Question:")[1].strip()
        questions.append(question)

    return questions

# Streamlit app interface
st.title("PDF to Question Generator")

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

if uploaded_file is not None:
    st.write("Extracting text from the PDF...")
    text = extract_text_from_pdf(uploaded_file)
    st.write("Text extracted successfully!")
    st.write("Extracted Text:", text)

    st.write("Generating questions...")
    num_questions = st.number_input("Number of questions to generate", min_value=1, max_value=20, value=5, step=1, format="%d")
    questions = generate_questions(text, num_questions)

    st.write("Generated Questions:")
    for idx, question in enumerate(questions):
        st.write(f"{idx+1}. {question}")