app.py

import spaces
import gradio as gr
from transformers import pipeline
from PIL import Image

# Load model and processor
pipe = pipeline("image-to-text", model="google/pix2struct-screen2words-large")
# Define the function
@spaces.GPU
def describe_ui(image):
    outputs = pipe(image, text="describe this image.")
    return outputs[0]['generated_text']

# Launch the Gradio interface
gr.Interface(
    fn=describe_ui,
    inputs=gr.Image(type="pil"),
    outputs="text",
    title="UI Screen Describer (Pix2Struct)",
    description="Upload a screenshot or UI image and get an automatic description powered by Google’s Pix2Struct model."
).launch()