app.py

import gradio as gr
import tensorflow as tf
from tensorflow.keras.preprocessing import image
import matplotlib.pyplot as plt
import numpy as np
model = tf.keras.models.load_model('dogcat_model_bak.h5')
def image_classifier(img):
    img1 = image.load_img(str(img), target_size=(64, 64))
    img1 = image.img_to_array(img1)
    img1 = img1/255
    img1 = np.expand_dims(img1, axis=0)
    res = model.predict(img, batch_size=None,steps=1)
    if(res[:,:]>0.5):
      value ='Dog :%1.2f'%(prediction[0,0])
    else:
      value ='Cat :%1.2f'%(1.0-prediction[0,0])
    return value

demo = gr.Interface(fn=image_classifier, inputs="image", outputs="label")
demo.launch()