app.py

import gradio as gr
import numpy as np 
from PIL import Image, ImageOps
import tensorflow as tf

#Load the model 
model = tf.keras.models.load_model("cnn_model.h5")

def predict(image_array):
    try:
        print("Function called!")

        # Check for blank or None input
        if image_array is None or np.sum(image_array) == 0:
            return "Please draw a digit."

        # Convert to PIL image and process
        image = Image.fromarray(image_array.astype("uint8"), mode="L")
        image = ImageOps.invert(image).resize((28, 28))

        # Normalize and reshape for model input
        image_array = np.array(image).astype("float32") / 255.0
        image_array = image_array.reshape(1, 28, 28, 1)

        # Predict
        logits = model.predict(image_array)
        prediction = int(np.argmax(logits))
        confidence = float(tf.nn.softmax(logits)[0][prediction])

        return f"Digit: {prediction} (confidence: {confidence:.2%})"

    except Exception as err:
        return f"💥 Runtime error: {str(err)}"


gr.Sketchpad(image_mode="L", canvas_size=(200, 200), type="numpy")
gr.Interface(
    fn=predict,
    inputs=gr.Sketchpad(image_mode="L", canvas_size=(200, 200)),
    outputs="text",
    title="Digit Classifier",
).launch()