import gradio as gr # import torch # from torch import autocast # from diffusers import StableDiffusionPipeline from datasets import load_dataset from PIL import Image from io import BytesIO # import base64 # import re import os import requests import json import base64 # from urllib import parse from share_btn import community_icon_html, loading_icon_html, share_js is_gpu_busy = False def infer(prompt, n_samples, steps, scale, seed): global is_gpu_busy # generator = torch.Generator(device=device).manual_seed(seed) # print("Is GPU busy? ", is_gpu_busy) images = [] # if(not is_gpu_busy): # is_gpu_busy = True # images_list = pipe( # [prompt] * samples, # num_inference_steps=steps, # guidance_scale=scale, # generator=generator, # ) # is_gpu_busy = False # safe_image = Image.open(r"unsafe.png") # for i, image in enumerate(images_list["sample"]): # if(images_list["nsfw_content_detected"][i]): # images.append(safe_image) # else: # images.append(image) # else: url = os.getenv('BACKEND_URL') response = requests.get(url.format(prompt, int(n_samples), max(50,int(steps)), f'{scale:.1f}', int(seed))) #response = requests.get(url.format('a%20naked%20girl', 2, 50, 7.5, 2)) data = json.load(BytesIO(response.content)) if 'output' not in data: raise gr.Error("Although safety guidance is enabled, potential unsafe content found. Please try again with different seed.") else: for image in data['output']['choices']: im = Image.open(BytesIO(base64.b64decode(image['image_base64']))) images.append(im) # payload = {'prompt': prompt} # images_request = requests.post(url, json=payload) # for image in images_request.json()["output"]['choices']: # image_b64 = (f"data:image/jpeg;base64,{image['image_base64']}") # images.append(image_b64) return images css = """ .gradio-container { font-family: 'IBM Plex Sans', sans-serif; } .gr-button { color: white; border-color: #3a669bff; background: #3a669bff; } input[type='range'] { accent-color: #3a669bff; } .dark input[type='range'] { accent-color: #3a669bff; } .container { max-width: 730px; margin: auto; padding-top: 1.5rem; } #gallery { min-height: 22rem; margin-bottom: 15px; margin-left: auto; margin-right: auto; border-bottom-right-radius: .5rem !important; border-bottom-left-radius: .5rem !important; } #gallery>div>.h-full { min-height: 20rem; } .details:hover { text-decoration: underline; } .gr-button { white-space: nowrap; } .gr-button:focus { border-color: rgb(147 197 253 / var(--tw-border-opacity)); outline: none; box-shadow: var(--tw-ring-offset-shadow), var(--tw-ring-shadow), var(--tw-shadow, 0 0 #0000); --tw-border-opacity: 1; --tw-ring-offset-shadow: var(--tw-ring-inset) 0 0 0 var(--tw-ring-offset-width) var(--tw-ring-offset-color); --tw-ring-shadow: var(--tw-ring-inset) 0 0 0 calc(3px var(--tw-ring-offset-width)) var(--tw-ring-color); --tw-ring-color: rgb(191 219 254 / var(--tw-ring-opacity)); --tw-ring-opacity: .5; } #advanced-btn { font-size: .7rem !important; line-height: 19px; margin-top: 12px; margin-bottom: 12px; padding: 2px 8px; border-radius: 14px !important; } #advanced-options { display: none; margin-bottom: 20px; } .footer { margin-bottom: 45px; margin-top: 35px; text-align: center; border-bottom: 1px solid #e5e5e5; } .footer>p { font-size: .8rem; display: inline-block; padding: 0 10px; transform: translateY(10px); background: white; } .dark .footer { border-color: #303030; } .dark .footer>p { background: #0b0f19; } .acknowledgments h4{ margin: 1.25em 0 .25em 0; font-weight: bold; font-size: 115%; } #container-advanced-btns{ display: flex; flex-wrap: wrap; justify-content: space-between; align-items: center; } .animate-spin { animation: spin 1s linear infinite; } @keyframes spin { from { transform: rotate(0deg); } to { transform: rotate(360deg); } } #share-btn-container { display: flex; padding-left: 0.5rem !important; padding-right: 0.5rem !important; background-color: #3a669bff; justify-content: center; align-items: center; border-radius: 9999px !important; width: 13rem; } #share-btn { all: initial; color: #ffffff;font-weight: 600; cursor:pointer; font-family: 'IBM Plex Sans', sans-serif; margin-left: 0.5rem !important; padding-top: 0.25rem !important; padding-bottom: 0.25rem !important; } #share-btn * { all: unset; } .gr-form{ flex: 1 1 50%; border-top-right-radius: 0; border-bottom-right-radius: 0; } #prompt-container{ gap: 0; } """ block = gr.Blocks(css=css) examples = [ [ 'a gorgeous female photo, professionally retouched, soft lighting, torso, legs, feet, realistic, smooth face, perfect eyes, !! wide angle!!', 2, 50, 7.5, 1, ], [ 'portrait of girl with smokey eyes makeup in abandoned hotel, grange clothes, redshift, wide high angle coloured polaroid photograph with flash, kodak film, hyper real, stunning moody cinematography, with anamorphic lenses, by maripol, fallen angels by wong kar - wai, style of suspiria and neon demon and children from bahnhof zoo, detailed ', 2, 50, 9, 364629028, ], [ 'padme amidala taking a bath artwork, safe for work, no nudity', 2, 50, 7.5, 0, ], #[ # 'portrait of Sickly diseased dying Samurai warrior, sun shining, photo realistic illustration by greg rutkowski, thomas kindkade, alphonse mucha, loish, norman rockwell.', # 2, # 50, # 7.5, # 1, #], [ 'a photograph by vanessa beecroft', 2, 50, 7.5, 445713656, ], [ 'a disturbing horror photograph of a beautiful crying woman with flowers and fungus growing out of her head m, intricate, hyperrealism, sharp focus, cinematography, highly detailed, octane render, digital horror artwork, matte, photography by professional photographe', 2, 50, 7.5, 1, ], ] with block: gr.HTML( """
Safe Stable Diffusion extends Stable Diffusion with safety guidance. In the case of NSFW images it returns the closest non-NSFW images instead of a black square. Details can be found in the Safe Latent Diffusion: Mitigating Inappropriate Degeneration in Diffusion Models paper.