app.py

import gradio as gr
import torch
import modin.pandas as pd
from diffusers import DiffusionPipeline 

device = "cuda" if torch.cuda.is_available() else "cpu"
if torch.cuda.is_available():
    PYTORCH_CUDA_ALLOC_CONF={'max_split_size_mb': 6000}
    torch.cuda.max_memory_allocated(device=device)
    torch.cuda.empty_cache()
    pipe = DiffusionPipeline.from_pretrained("stabilityai/sdxl-turbo", torch_dtype=torch.float16, variant="fp16", use_safetensors=True)
    pipe.enable_xformers_memory_efficient_attention()
    pipe = pipe.to(device)
    pipe.unet = torch.compile(pipe.unet, mode="reduce-overhead", fullgraph=True)
    torch.cuda.empty_cache()
    refiner = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-refiner-1.0", use_safetensors=True, torch_dtype=torch.float16, variant="fp16")
    refiner.enable_xformers_memory_efficient_attention()
    refiner.enable_sequential_cpu_offload()
    refiner.unet = torch.compile(refiner.unet, mode="reduce-overhead", fullgraph=True)
else: 
    pipe = DiffusionPipeline.from_pretrained("stabilityai/sdxl-turbo", use_safetensors=True)
    pipe = pipe.to(device)
    pipe.unet = torch.compile(pipe.unet, mode="reduce-overhead", fullgraph=True)
    refiner = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-refiner-1.0", use_safetensors=True)
    refiner = refiner.to(device)
    refiner.unet = torch.compile(refiner.unet, mode="reduce-overhead", fullgraph=True)

def genie (prompt, negative_prompt, height, width, scale, steps, seed, prompt_2, negative_prompt_2, high_noise_frac):
    generator = np.random.seed(0) if seed == 0 else torch.manual_seed(seed)
    int_image = pipe(prompt, prompt_2=prompt_2, height=height, width=width, num_inference_steps=steps, num_images_per_prompt=1, generator=generator, output_type="latent").images 
    image = refiner(prompt=prompt, prompt_2=prompt_2, negative_prompt=negative_prompt, negative_prompt_2=negative_prompt_2, image=int_image, denoising_start=high_noise_frac).images[0]   
    return image
    
gr.Interface(fn=genie, inputs=[gr.Textbox(label='What you want the AI to generate. 77 Token Limit.'), 
    gr.Textbox(label='What you Do Not want the AI to generate.'), 
    gr.Slider(512, 1024, 768, step=128, label='Height'),
    gr.Slider(512, 1024, 768, step=128, label='Width'),
    gr.Slider(1, 15, 10, label='Guidance Scale'), 
    gr.Slider(1, maximum=5, value=2, step=1, label='Number of Iterations'), 
    gr.Slider(minimum=1, step=1, maximum=999999999999999999, randomize=True),
    gr.Textbox(label='Embedded Prompt'),
    gr.Textbox(label='Embedded Negative Prompt'),
    gr.Slider(minimum=.7, maximum=.99, value=.95, step=.01, label='Refiner Denoise Start %')],
    outputs='image', 
    title="Stable Diffusion XL 1.0 CPU or GPU", 
    description="SDXL 1.0 CPU or GPU. Currently running on CPU. <br><br><b>WARNING:</b> Extremely Slow. 65s/Iteration. Expect 25-50mins an image for 25-50 iterations respectively. This model is capable of producing NSFW (Softcore) images.", 
    article = "If You Enjoyed this Demo and would like to Donate, you can send to any of these Wallets. <br>BTC: bc1qzdm9j73mj8ucwwtsjx4x4ylyfvr6kp7svzjn84 <br>3LWRoKYx6bCLnUrKEdnPo3FCSPQUSFDjFP <br>DOGE: DK6LRc4gfefdCTRk9xPD239N31jh9GjKez <br>SHIB (BEP20): 0xbE8f2f3B71DFEB84E5F7E3aae1909d60658aB891 <br>PayPal: https://www.paypal.me/ManjushriBodhisattva <br>ETH: 0xbE8f2f3B71DFEB84E5F7E3aae1909d60658aB891 <br>Code Monkey: <a href=\"https://huggingface.co/Manjushri\">Manjushri</a>").launch(debug=True, max_threads=80)