app.py

import gradio as gr
from transformers import pipeline
# Initialize the question-answering pipeline
qa_model = pipeline('question-answering', model='deepset/roberta-base-squad2')
def get_answer(context, question):
    # Check if both inputs are provided
    if not context.strip() or not question.strip():
        return "Error: Please provide both context and a question."
    # Run the model to find the answer
    result = qa_model(question=question, context=context)
    # Extract the answer text
    answer_text = result['answer']
    # Calculate the confidence score
    confidence = round(result['score'] * 100, 1)
    # Return formatted string with answer and probability
    return f"{answer_text} (Confidence: {confidence}%)"
# Create Gradio blocks interface
with gr.Blocks(title="QA Assistant", theme=gr.themes.Soft()) as demo:
    # Add main header
    gr.Markdown("# AI Assistant: Question Answering")
    # Add application description
    gr.Markdown("Provide a context text and ask a question about it. The AI will analyze the text and extract the exact answer.")
    # Create layout with two columns
    with gr.Row():
        with gr.Column(scale=2):
            # Add input box for the context text
            input_context = gr.Textbox(
                label="Context (Text)", 
                lines=8, 
                placeholder="Paste an article or paragraph in English here..."
            )
            # Add input box for the question
            input_question = gr.Textbox(
                label="Your Question", 
                lines=2, 
                placeholder="What is this text about?"
            )
            # Add action button
            btn_answer = gr.Button("Find Answer", variant='primary')
        with gr.Column(scale=1):
            # Add output box for the answer
            output_answer = gr.Textbox(
                label="AI Answer", 
                lines=4, 
                interactive=False
            )
    # Link button click to the answering function
    btn_answer.click(
        fn=get_answer, 
        inputs=[input_context, input_question], 
        outputs=output_answer
    )
    # Add examples for quick testing
    gr.Examples(
        examples=[
            [
                "Python was created by Guido van Rossum and first released in 1991. It is a widely used high-level programming language.", 
                "Who created Python?"
            ],
            [
                "The James Webb Space Telescope (JWST) is a space telescope designed primarily to conduct infrared astronomy. It was launched in December 2021.", 
                "When was the telescope launched?"
            ]
        ],
        inputs=[input_context, input_question]
    )
    
if __name__ == '__main__':
    demo.launch()