import gradio as gr import torch from diffusers import AutoPipelineForInpainting import diffusers from share_btn import community_icon_html, loading_icon_html, share_js from sdxl import sdxl_diffusion_loop from sdxl_models import SDXLUNet, SDXLVae, SDXLControlNetPreEncodedControlnetCond import torchvision.transforms.functional as TF from diffusion import make_sigmas, set_with_tqdm from huggingface_hub import hf_hub_download import gc set_with_tqdm(True) pipe = AutoPipelineForInpainting.from_pretrained("diffusers/stable-diffusion-xl-1.0-inpainting-0.1", torch_dtype=torch.float16, variant="fp16") pipe.text_encoder.to("cuda") pipe.text_encoder_2.to("cuda") comparing_unet = SDXLUNet.load(hf_hub_download("stabilityai/stable-diffusion-xl-base-1.0", "unet/diffusion_pytorch_model.fp16.safetensors")) comparing_vae = SDXLVae.load(hf_hub_download("madebyollin/sdxl-vae-fp16-fix", "diffusion_pytorch_model.safetensors")) comparing_vae.to(torch.float16) comparing_controlnet = SDXLControlNetPreEncodedControlnetCond.load(hf_hub_download("williamberman/sdxl_controlnet_inpainting", "sdxl_controlnet_inpaint_pre_encoded_controlnet_cond_checkpoint_200000.safetensors")) comparing_controlnet.to(torch.float16) gc.collect() torch.cuda.empty_cache() def read_content(file_path: str) -> str: """read the content of target file """ with open(file_path, 'r', encoding='utf-8') as f: content = f.read() return content def predict(dict, prompt="", negative_prompt="", guidance_scale=7.5, steps=20, strength=1.0, scheduler="EulerDiscreteScheduler"): if negative_prompt == "": negative_prompt = None scheduler_class_name = scheduler.split("-")[0] add_kwargs = {} if len(scheduler.split("-")) > 1: add_kwargs["use_karras"] = True if len(scheduler.split("-")) > 2: add_kwargs["algorithm_type"] = "sde-dpmsolver++" scheduler = getattr(diffusers, scheduler_class_name) pipe.scheduler = scheduler.from_pretrained("stabilityai/stable-diffusion-xl-base-1.0", subfolder="scheduler", **add_kwargs) init_image = dict["image"].convert("RGB").resize((1024, 1024)) mask = dict["mask"].convert("RGB").resize((1024, 1024)) pipe.vae.to('cuda') pipe.unet.to('cuda') output = pipe(prompt = prompt, negative_prompt=negative_prompt, image=init_image, mask_image=mask, guidance_scale=guidance_scale, num_inference_steps=int(steps), strength=strength) pipe.vae.to('cpu') pipe.unet.to('cpu') gc.collect() torch.cuda.empty_cache() comparing_vae.to('cuda') comparing_unet.to('cuda') comparing_controlnet.to('cuda') image = TF.to_tensor(dict["image"].convert("RGB").resize((1024, 1024))) mask = TF.to_tensor(dict["mask"].convert("L").resize((1024, 1024))) image = image * (mask < 0.5) image = TF.normalize(image, [0.5], [0.5]) image = comparing_vae.encode(image[None, :, :, :].to(dtype=comparing_vae.dtype, device=comparing_vae.device)).to(dtype=comparing_controlnet.dtype, device=comparing_controlnet.device) mask = TF.resize(mask, (1024 // 8, 1024 // 8))[None, :, :, :].to(dtype=image.dtype, device=image.device) image = torch.concat((image, mask), dim=1) sigmas = make_sigmas(device=comparing_unet.device).to(dtype=comparing_unet.dtype) timesteps = torch.linspace(0, sigmas.numel() - 1, int(steps), dtype=torch.long, device=comparing_unet.device) out = sdxl_diffusion_loop( prompts=prompt, negative_prompts=negative_prompt, images=image, guidance_scale=guidance_scale, sigmas=sigmas, timesteps=timesteps, text_encoder_one=pipe.text_encoder, text_encoder_two=pipe.text_encoder_2, unet=comparing_unet, controlnet=comparing_controlnet ) comparing_unet.to('cpu') comparing_controlnet.to('cpu') gc.collect() torch.cuda.empty_cache() out = comparing_vae.output_tensor_to_pil(comparing_vae.decode(out)) comparing_vae.to('cpu') gc.collect() torch.cuda.empty_cache() return output.images[0], out[0], gr.update(visible=True) css = ''' .gradio-container{max-width: 1100px !important} #image_upload{min-height:400px} #image_upload [data-testid="image"], #image_upload [data-testid="image"] > div{min-height: 400px} #mask_radio .gr-form{background:transparent; border: none} #word_mask{margin-top: .75em !important} #word_mask textarea:disabled{opacity: 0.3} .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%} #image_upload .touch-none{display: flex} @keyframes spin { from { transform: rotate(0deg); } to { transform: rotate(360deg); } } #share-btn-container {padding-left: 0.5rem !important; padding-right: 0.5rem !important; background-color: #000000; justify-content: center; align-items: center; border-radius: 9999px !important; max-width: 13rem; margin-left: auto;} div#share-btn-container > div {flex-direction: row;background: black;align-items: center} #share-btn-container:hover {background-color: #060606} #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: 0.5rem !important; padding-bottom: 0.5rem !important;right:0;} #share-btn * {all: unset} #share-btn-container div:nth-child(-n+2){width: auto !important;min-height: 0px !important;} #share-btn-container .wrap {display: none !important} #share-btn-container.hidden {display: none!important} #prompt input{width: calc(100% - 160px);border-top-right-radius: 0px;border-bottom-right-radius: 0px;} #run_button{position:absolute;margin-top: 11px;right: 0;margin-right: 0.8em;border-bottom-left-radius: 0px; border-top-left-radius: 0px;} #prompt-container{margin-top:-18px;} #prompt-container .form{border-top-left-radius: 0;border-top-right-radius: 0} #image_upload{border-bottom-left-radius: 0px;border-bottom-right-radius: 0px} ''' image_blocks = gr.Blocks(css=css, elem_id="total-container") with image_blocks as demo: gr.HTML(read_content("header.html")) with gr.Row(): with gr.Column(): image = gr.Image(source='upload', tool='sketch', elem_id="image_upload", type="pil", label="Upload",height=400) with gr.Row(elem_id="prompt-container", mobile_collapse=False, equal_height=True): with gr.Row(): prompt = gr.Textbox(placeholder="Your prompt (what you want in place of what is erased)", show_label=False, elem_id="prompt") btn = gr.Button("Inpaint!", elem_id="run_button") with gr.Accordion(label="Advanced Settings", open=False): with gr.Row(mobile_collapse=False, equal_height=True): guidance_scale = gr.Number(value=7.5, minimum=1.0, maximum=20.0, step=0.1, label="guidance_scale") steps = gr.Number(value=20, minimum=1, maximum=1000, step=1, label="steps") strength = gr.Number(value=0.99, minimum=0.01, maximum=1.0, step=0.01, label="strength") negative_prompt = gr.Textbox(label="negative_prompt", placeholder="Your negative prompt", info="what you don't want to see in the image") with gr.Row(mobile_collapse=False, equal_height=True): schedulers = ["DEISMultistepScheduler", "HeunDiscreteScheduler", "EulerDiscreteScheduler", "DPMSolverMultistepScheduler", "DPMSolverMultistepScheduler-Karras", "DPMSolverMultistepScheduler-Karras-SDE"] scheduler = gr.Dropdown(label="Schedulers", choices=schedulers, value="EulerDiscreteScheduler") with gr.Column(): image_out = gr.Image(label="Output diffusers full finetune 0.1", elem_id="output-img", height=400) image_out_comparing = gr.Image(label="Output controlnet + vae", elem_id="output-img-comparing", height=400) with gr.Group(elem_id="share-btn-container", visible=False) as share_btn_container: community_icon = gr.HTML(community_icon_html) loading_icon = gr.HTML(loading_icon_html) share_button = gr.Button("Share to community", elem_id="share-btn",visible=True) btn.click(fn=predict, inputs=[image, prompt, negative_prompt, guidance_scale, steps, strength, scheduler], outputs=[image_out, image_out_comparing, share_btn_container], api_name='run') prompt.submit(fn=predict, inputs=[image, prompt, negative_prompt, guidance_scale, steps, strength, scheduler], outputs=[image_out, image_out_comparing, share_btn_container]) share_button.click(None, [], [], _js=share_js) gr.Examples( examples=[ ["./imgs/aaa (8).png"], ["./imgs/download (1).jpeg"], ["./imgs/0_oE0mLhfhtS_3Nfm2.png"], ["./imgs/02_HubertyBlog-1-1024x1024.jpg"], ["./imgs/jdn_jacques_de_nuce-1024x1024.jpg"], ["./imgs/c4ca473acde04280d44128ad8ee09e8a.jpg"], ["./imgs/canam-electric-motorcycles-scaled.jpg"], ["./imgs/e8717ce80b394d1b9a610d04a1decd3a.jpeg"], ["./imgs/Nature___Mountains_Big_Mountain_018453_31.jpg"], ["./imgs/Multible-sharing-room_ccexpress-2-1024x1024.jpeg"], ], fn=predict, inputs=[image], cache_examples=False, ) gr.HTML( """ """ ) image_blocks.queue(max_size=25).launch()