ask2pdf.py

import pdfplumber
from transformers import pipeline
import gradio as gr

# Load the pre-trained question-answering model
qa_pipeline = pipeline("question-answering", model="distilbert-base-uncased-distilled-squad")

# Function to extract text from the PDF
def extract_text_from_pdf(pdf_path):
    text = ""
    with pdfplumber.open(pdf_path) as pdf:
        for page in pdf.pages:
            text += page.extract_text()
    return text

# Function to handle question-answering
def qa_interface(pdf, question):
    if not pdf:
        return "Please upload a PDF file."
    if not question.strip():
        return "Please enter a question."
    try:
        context = extract_text_from_pdf(pdf.name)
        if len(context.strip()) == 0:
            return "The PDF appears to be empty or text could not be extracted. Please try another file."
        result = qa_pipeline(question=question, context=context)
        return result["answer"]
    except Exception as e:
        return f"An error occurred: {str(e)}"

# Gradio Interface
demo = gr.Interface(
    fn=qa_interface,
    inputs=[
        gr.File(label="Upload PDF", file_types=[".pdf"]),
        gr.Textbox(label="Enter your question", placeholder="What do you want to know about the PDF?"),
    ],
    outputs=gr.Textbox(label="Answer"),
    title="PDF Question Answering by Aitude.com",
    description="""
    Welcome to the PDF Question Answering demo powered by Aitude.com. 
    - Upload a PDF document (ensure it contains selectable text, as scanned PDFs are not supported).
    - Enter a question based on the content of the uploaded PDF.
    - The model will extract the relevant answer from the PDF.
    
    This demo is ideal for querying reports, articles, and other text-based documents effortlessly.
    """,
    examples=[
        ["aitude.pdf", "What is AITUDE's core area of expertise?"],
        ["aitude.pdf", "How does AITUDE leverage natural language processing (NLP) models?"],
        ["aitude.pdf", "What is Generative AI and how is AITUDE involved in this field?"],
        ["aitude.pdf", "What are some potential applications of AITUDE's advancements in Generative AI?"],
        ["aitude.pdf", "How has Sandeep Kumar Mishra's leadership contributed to AITUDE's success?"],
    ],
)

if __name__ == "__main__":
    demo.launch()