import gradio as gr from PIL import Image from io import BytesIO import torch import os from diffusers import DiffusionPipeline, DDIMScheduler MY_SECRET_TOKEN=os.environ.get('HF_TOKEN_SD') has_cuda = torch.cuda.is_available() device = torch.device('cpu' if not has_cuda else 'cuda:0') pipe = DiffusionPipeline.from_pretrained( "CompVis/stable-diffusion-v1-4", safety_checker=None, use_auth_token=MY_SECRET_TOKEN, custom_pipeline="imagic_stable_diffusion", scheduler = DDIMScheduler(beta_start=0.00085, beta_end=0.012, beta_schedule="scaled_linear", clip_sample=False, set_alpha_to_one=False) ).to(device) generator = torch.Generator("cuda").manual_seed(0) def infer(prompt, init_image): init_image = Image.open(init_image).convert("RGB") init_image = init_image.resize((512, 512)) res = pipe.train( prompt, init_image, guidance_scale=7.5, num_inference_steps=50, generator=generator) res = pipe(alpha=1) return res.images[0] title = """
Text-Based Real Image Editing with Diffusion Models
This pipeline aims to implement this paper to Stable Diffusion, allowing for real-world image editing.