app.py

import gradio as gr
from diffusers import AutoPipelineForInpainting, AutoencoderKL
import torch
from PIL import Image, ImageOps

vae = AutoencoderKL.from_pretrained("madebyollin/sdxl-vae-fp16-fix", torch_dtype=torch.float16)
pipeline = AutoPipelineForInpainting.from_pretrained("diffusers/stable-diffusion-xl-1.0-inpainting-0.1", vae=vae, torch_dtype=torch.float16, variant="fp16", use_safetensors=True).to("cuda")


def divisible_by_8(image):
    width, height = image.size
    
    # Calculate the new width and height that are divisible by 8
    new_width = (width // 8) * 8
    new_height = (height // 8) * 8
    
    # Resize the image
    resized_image = image.resize((new_width, new_height))
    
    return resized_image


def generate(image_editor, prompt, neg_prompt, strength, guidance):
    image = image_editor['background'].convert('RGB')
    image.thumbnail((1024, 1024))
    image = divisible_by_8(image)

    layer = image_editor["layers"][0].resize(image.size)
    
    mask = Image.new("RGBA", image.size, "WHITE") 
    mask.paste(layer, (0, 0), layer)
    mask = ImageOps.invert(mask.convert('L'))
    

    final_image = pipeline(prompt=prompt, 
                           image=image, 
                           mask_image=mask).images[0]
                           #width=image.width, 
                           #height=image.height, 
                           #num_inference_steps=50, 
                           #strength=strength, 
                           #guidance_scale=guidance).images[0]

    return image_editor, image, mask, final_image

with gr.Blocks() as demo:
    gr.Markdown("""
    # Inpainting Sketch Pad
    by [Tony Assi](https://www.tonyassi.com/)
    """)
    
    with gr.Row():
        with gr.Column():
            sketch_pad = gr.ImageMask(type='pil', label='Inpaint')
            prompt = gr.Textbox()
            generate_button = gr.Button("Generate")
        with gr.Column():
            version_gallery = gr.Gallery(label="Versions")
            restore_button = gr.Button("Restore Version")

    with gr.Accordion("Advanced Settings", open=False):
        neg_prompt = gr.Textbox(label='Negative Prompt', value='ugly, deformed, nsfw')
        strength_slider = gr.Slider(0.0, 1.0, value=1.0, label="Strength")
        guidance_slider = gr.Slider(1.0, 15.0, value=7.5, label="Guidance")
        
    with gr.Row():
        out1 = gr.Image(format='png')
        out2 = gr.Image(format='png')
        out3 = gr.Image(format='png')
        
    generate_button.click(fn=generate, inputs=[sketch_pad,prompt, neg_prompt, strength_slider, guidance_slider], outputs=[sketch_pad, out1, out2, out3])

demo.launch()