import gradio as gr
import tensorflow as tf
from huggingface_hub import from_pretrained_keras
import numpy as np

from huggingface_hub import hf_hub_download
from keras.layers import TFSMLayer
import numpy as np

# Download the model
model_path = hf_hub_download("keras-io/mobile-vit-xxs")

# Load the model using TFSMLayer
model = TFSMLayer(model_path, call_endpoint='serving_default')

''' model = from_pretrained_keras("keras-io/mobile-vit-xxs") 
'''

classes=['dandelion','daisy','tulips','sunflower','rose']
image_size = 256

def classify_images(image):
  image = tf.convert_to_tensor(image)
  image = tf.image.resize(image, (image_size, image_size))
  image = tf.expand_dims(image,axis=0)
  prediction = model.predict(image)
  prediction = tf.squeeze(tf.round(prediction))
  text_output = str(f'{classes[(np.argmax(prediction))]}!')
  return text_output
  
i = gr.inputs.Image()
o = gr.outputs.Textbox()

examples = [["./examples/tulip.jpg"], ["./examples/daisy.jpg"], ["./examples/dandelion.jpg"], ["./examples/rose.jpg"], ["./examples/sunflower.jpg"]]
title = "Flower Recognition Using Transfer Learning"
description = "Upload an image or Select from the examples below to classify flowers: "

article = "<div style='text-align: center;'><a href='https://www.linkedin.com/in/suryakiran-mg/' target='_blank'> Space by Suryakiran </a></div>"
gr.Interface(classify_images, i, o, allow_flagging=False, analytics_enabled=False,
  title=title, examples=examples, description=description, article=article).launch(enable_queue=True)