app.py

import numpy as np
import tensorflow as tf
import cv2
import gradio as gr
from tensorflow import keras

model = keras.models.load_model('model_CNN.h5')

class_mapping = {1: 'Собака', 2: 'Кінь', 3: 'Слон', 4:'Метелик',
                 5: 'Курка', 6: 'Кіт', 7:'Корова', 8: 'Вівця',
                 9: 'Павук', 10: 'Білка'
                }

  # Створення функції для передбачення тварини
def predict_image(image):
    image = cv2.resize(image, (224, 224))
    image = np.asarray(image)
    image = image.astype('float32') / 255.0
    predictions = model.predict(np.expand_dims(image, axis=0))[0]
    prediction = {}
    for index, probability in enumerate(predictions) :
      prediction[class_mapping[index+1]] = float(round(probability, 3))
      print(prediction)
    return prediction

demo = gr.Blocks()
# Створення інтерфейсу Gradio
with demo:
    gr.Markdown("What animal is in the picture")
    with gr.Tab("Predict image"):
        image_input = gr.Image(label="Upload image")
        output = gr.Label(label="Animal predicted by neural network")
        image_button = gr.Button("Predict")
        image_button.click(predict_image, inputs=image_input, outputs=output)

demo.launch()