app.py

import gradio as gr
import transforms
import requests
import io
from PIL import Image
from transformers import pipeline
from torchvision import transforms

title = "Fine Tuned SD Model - Authoral stylization"
description = "Generate images trained in an authoral illustration model."
article = "<p style='text-align: center'><a href='https://news.machinelearning.sg/posts/beautiful_profile_pics_remove_background_image_with_deeplabv3/'>Blog</a> | <a href='https://github.com/eugenesiow/practical-ml'>Github Repo</a></p>"


gr.Interface.load(
  "spaces/Cacau/heart-of-painting",

demo = gr.Interface(
    fn=greet,
    inputs=gr.Textbox(lines=2, placeholder="Name Here..."),
    outputs="text",
)


with gr.Blocks(theme=gr.themes.Glass()) as demo:
   inputs=[gr.Textbox(label="Prompt", source="input box")]
output=[gr.Ima]
).launch()
    
def query(payload):
	response = requests.post(API_URL, headers=headers, json=payload)
	return response.content
image_bytes = query({
	"inputs":gr.Textbox(lines=2, placeholder="Your prompt here..."),    
})

# You can access the image with PIL.Image for example
import io
from PIL import Image
image = Image.open(io.BytesIO(image_bytes))

return "This is your generated image:" + image "**Save it in your files!"
demo.launch()