app.py

import gradio as gr
import spaces
import os
os.environ["HF_HUB_ENABLE_HF_TRANSFER"] = "1"

import torch

#Hack for ZeroGPU
torch.jit.script = lambda f: f
####

from omni_zero import OmniZeroCouple

omni_zero = OmniZeroCouple(
    base_model="frankjoshua/albedobaseXL_v13",
    device="cuda",
)

@spaces.GPU()
def generate(
    base_image="https://cdn-prod.styleof.com/inferences/cm1ho5cjl14nh14jec6phg2h8/i6k59e7gpsr45ufc7l8kun0g-medium.jpeg",
    style_image="https://cdn-prod.styleof.com/inferences/cm1ho5cjl14nh14jec6phg2h8/i6k59e7gpsr45ufc7l8kun0g-medium.jpeg",
    identity_image_1="https://cdn-prod.styleof.com/inferences/cm1hp4lea14oz14jeoghnex7g/dlgc5xwo0qzey7qaixy45i1o-medium.jpeg",
    identity_image_2="https://cdn-prod.styleof.com/inferences/cm1ho69ha14np14jesnusqiep/mp3aaktzqz20ujco5i3bi5s1-medium.jpeg",
    seed=42,
    prompt="Cinematic still photo of a couple. emotional, harmonious, vignette, 4k epic detailed, shot on kodak, 35mm photo, sharp focus, high budget, cinemascope, moody, epic, gorgeous, film grain, grainy",
    negative_prompt="anime, cartoon, graphic, (blur, blurry, bokeh), text, painting, crayon, graphite, abstract, glitch, deformed, mutated, ugly, disfigured",
    guidance_scale=3.0,
    number_of_images=1,
    number_of_steps=10,
    base_image_strength=0.3,
    style_image_strength=1.0,
    identity_image_strength_1=1.0,
    identity_image_strength_2=1.0,
    depth_image=None,
    depth_image_strength=0.2,
    mask_guidance_start=0.0,
    mask_guidance_end=1.0,
    progress=gr.Progress(track_tqdm=True)
):
    images = omni_zero.generate(
        seed=seed,
        prompt=prompt,
        negative_prompt=negative_prompt,
        guidance_scale=guidance_scale,
        number_of_images=number_of_images,
        number_of_steps=number_of_steps,
        base_image=base_image,
        base_image_strength=base_image_strength,
        style_image=style_image,
        style_image_strength=style_image_strength,
        identity_image_1=identity_image_1,
        identity_image_strength_1=identity_image_strength_1,
        identity_image_2=identity_image_2,
        identity_image_strength_2=identity_image_strength_2,
        depth_image=depth_image,
        depth_image_strength=depth_image_strength,
        mask_guidance_start=mask_guidance_start,
        mask_guidance_end=mask_guidance_end,
    )

    return images

#Move the components in the example fields outside so they are available when gr.Examples is instantiated
buy_me_a_coffee_button = """
[![Buy me a coffee](https://img.buymeacoffee.com/button-api/?text=Buy%20me%20a%20coffee&emoji=&slug=vk654cf2pv8&button_colour=BD5FFF&font_colour=ffffff&font_family=Bree&outline_colour=000000&coffee_colour=FFDD00)](https://www.buymeacoffee.com/vk654cf2pv8)
"""

with gr.Blocks() as demo:
    gr.Markdown("<h1 style='text-align: center'>Omni Zero Couples</h1>")
    gr.Markdown("<h4 style='text-align: center'>A diffusion pipeline for zero-shot stylized portrait creation [<a href='https://github.com/okaris/omni-zero-couples' target='_blank'>GitHub</a>]")#, [<a href='https://styleof.com/s/remix-yourself' target='_blank'>StyleOf Remix Yourself</a>]</h4>")
    gr.Markdown(buy_me_a_coffee_button)

    with gr.Row():
        with gr.Column():
            with gr.Row():
                prompt = gr.Textbox(label="Prompt", value="Cinematic still photo of a couple. emotional, harmonious, vignette, 4k epic detailed, shot on kodak, 35mm photo, sharp focus, high budget, cinemascope, moody, epic, gorgeous, film grain, grainy")
            with gr.Row():
                negative_prompt = gr.Textbox(label="Negative Prompt", value="anime, cartoon, graphic, (blur, blurry, bokeh), text, painting, crayon, graphite, abstract, glitch, deformed, mutated, ugly, disfigured")
            with gr.Row():
                with gr.Column(min_width=140):
                    with gr.Row():
                        base_image = gr.Image(label="Base Image")
                    with gr.Row():
                        base_image_strength = gr.Slider(label="Strength",step=0.01, minimum=0.0, maximum=1.0, value=1.0)
            #with gr.Row():
                with gr.Column(min_width=140):
                    with gr.Row():
                        style_image = gr.Image(label="Style Image")
                    with gr.Row():
                        style_image_strength = gr.Slider(label="Strength",step=0.01, minimum=0.0, maximum=1.0, value=1.0)
            with gr.Row():
                with gr.Column(min_width=140):
                    with gr.Row():
                        identity_image = gr.Image(label="Identity Image")
                    with gr.Row():
                        identity_image_strength = gr.Slider(label="Strenght",step=0.01, minimum=0.0, maximum=1.0, value=1.0)
                with gr.Column(min_width=140):
                    with gr.Row():
                        identity_image_2 = gr.Image(label="Identity Image 2")
                    with gr.Row():
                        identity_image_strength_2 = gr.Slider(label="Strenght",step=0.01, minimum=0.0, maximum=1.0, value=1.0)
            with gr.Accordion("Advanced options", open=False):      
                with gr.Row():
                    seed = gr.Slider(label="Seed",step=1, minimum=0, maximum=10000000, value=42)
                    number_of_images = gr.Slider(label="Number of Outputs",step=1, minimum=1, maximum=4, value=1)
                with gr.Row():
                    guidance_scale = gr.Slider(label="Guidance Scale",step=0.1, minimum=0.0, maximum=14.0, value=3.0)
                    number_of_steps = gr.Slider(label="Number of Steps",step=1, minimum=1, maximum=50, value=10)
                with gr.Row():
                    mask_guidance_start = gr.Slider(label="Mask Guidance Start",step=0.01, minimum=0.0, maximum=1.0, value=0.0)
                    mask_guidance_end = gr.Slider(label="Mask Guidance End",step=0.01, minimum=0.0, maximum=1.0, value=1.0)
            
        with gr.Column():
            with gr.Row():
                out = gr.Gallery(label="Output(s)")
            with gr.Row():
                # clear = gr.Button("Clear")
                submit = gr.Button("Generate")
        
                submit.click(generate, inputs=[
                    prompt,
                    base_image,
                    style_image,
                    identity_image,
                    identity_image_2,
                    seed,
                    negative_prompt,
                    guidance_scale,
                    number_of_images,
                    number_of_steps,
                    base_image_strength,
                    style_image_strength,
                    identity_image_strength,
                    identity_image_strength_2,
                    mask_guidance_start,
                    mask_guidance_end,
                    ],
                    outputs=[out]
                )
        # clear.click(lambda: None, None, chatbot, queue=False)
    gr.Examples(
        examples=[
            [
                "Cinematic still photo of a couple. emotional, harmonious, vignette, 4k epic detailed, shot on kodak, 35mm photo, sharp focus, high budget, cinemascope, moody, epic, gorgeous, film grain, grainy",
                "https://cdn-prod.styleof.com/inferences/cm1ho5cjl14nh14jec6phg2h8/i6k59e7gpsr45ufc7l8kun0g-medium.jpeg",
                "https://cdn-prod.styleof.com/inferences/cm1ho5cjl14nh14jec6phg2h8/i6k59e7gpsr45ufc7l8kun0g-medium.jpeg",
                "https://cdn-prod.styleof.com/inferences/cm1hp4lea14oz14jeoghnex7g/dlgc5xwo0qzey7qaixy45i1o-medium.jpeg",
                "https://cdn-prod.styleof.com/inferences/cm1ho69ha14np14jesnusqiep/mp3aaktzqz20ujco5i3bi5s1-medium.jpeg"
            ]
        ],
        inputs=[prompt, base_image, style_image, identity_image, identity_image_2],
        outputs=[out],
        fn=generate,
        cache_examples="lazy",
    )
if __name__ == "__main__":
    demo.launch()