import numpy as np import PIL import inspect import os from tqdm import tqdm from typing import Any, Callable, Dict, List, Optional, Tuple, Union import torch import torch.nn.functional as F from transformers import ( CLIPImageProcessor, CLIPTextModel, CLIPTextModelWithProjection, CLIPTokenizer, CLIPVisionModelWithProjection, ) from diffusers.schedulers import KarrasDiffusionSchedulers from diffusers.image_processor import PipelineImageInput from diffusers import ( AutoencoderKL, UNet2DConditionModel, StableDiffusionXLPipeline, DDIMScheduler, EulerDiscreteScheduler, ) from diffusers.utils import BaseOutput from diffusers.utils.torch_utils import randn_tensor from pytorch_wavelets import DWTForward, DWTInverse from torchvision.transforms import GaussianBlur # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.rescale_noise_cfg def rescale_noise_cfg(noise_cfg, noise_pred_text, guidance_rescale=0.0): """ Rescale `noise_cfg` according to `guidance_rescale`. Based on findings of [Common Diffusion Noise Schedules and Sample Steps are Flawed](https://arxiv.org/pdf/2305.08891.pdf). See Section 3.4 """ std_text = noise_pred_text.std(dim=list(range(1, noise_pred_text.ndim)), keepdim=True) std_cfg = noise_cfg.std(dim=list(range(1, noise_cfg.ndim)), keepdim=True) # rescale the results from guidance (fixes overexposure) noise_pred_rescaled = noise_cfg * (std_text / std_cfg) # mix with the original results from guidance by factor guidance_rescale to avoid "plain looking" images noise_cfg = guidance_rescale * noise_pred_rescaled + (1 - guidance_rescale) * noise_cfg return noise_cfg # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.retrieve_timesteps def retrieve_timesteps( scheduler, num_inference_steps: Optional[int] = None, device: Optional[Union[str, torch.device]] = None, timesteps: Optional[List[int]] = None, **kwargs, ): """ Calls the scheduler's `set_timesteps` method and retrieves timesteps from the scheduler after the call. Handles custom timesteps. Any kwargs will be supplied to `scheduler.set_timesteps`. Args: scheduler (`SchedulerMixin`): The scheduler to get timesteps from. num_inference_steps (`int`): The number of diffusion steps used when generating samples with a pre-trained model. If used, `timesteps` must be `None`. device (`str` or `torch.device`, *optional*): The device to which the timesteps should be moved to. If `None`, the timesteps are not moved. timesteps (`List[int]`, *optional*): Custom timesteps used to support arbitrary spacing between timesteps. If `None`, then the default timestep spacing strategy of the scheduler is used. If `timesteps` is passed, `num_inference_steps` must be `None`. Returns: `Tuple[torch.Tensor, int]`: A tuple where the first element is the timestep schedule from the scheduler and the second element is the number of inference steps. """ if timesteps is not None: accepts_timesteps = "timesteps" in set(inspect.signature(scheduler.set_timesteps).parameters.keys()) if not accepts_timesteps: raise ValueError( f"The current scheduler class {scheduler.__class__}'s `set_timesteps` does not support custom" f" timestep schedules. Please check whether you are using the correct scheduler." ) scheduler.set_timesteps(timesteps=timesteps, device=device, **kwargs) timesteps = scheduler.timesteps num_inference_steps = len(timesteps) else: scheduler.set_timesteps(num_inference_steps, device=device, **kwargs) timesteps = scheduler.timesteps return timesteps, num_inference_steps def gaussian_blur_image_sharpening(image, kernel_size=3, sigma=(0.1, 2.0), alpha=1): gaussian_blur = GaussianBlur(kernel_size=kernel_size, sigma=sigma) image_blurred = gaussian_blur(image) image_sharpened = (alpha + 1) * image - alpha * image_blurred return image_sharpened class DiffuseHighSDXLPipelineOutput(BaseOutput): """ Output class for Stable Diffusion pipelines. Args: images (`List[PIL.Image.Image]` or `np.ndarray`) List of denoised PIL images of length `batch_size` or numpy array of shape `(batch_size, height, width, num_channels)`. PIL images or numpy array present the denoised images of the diffusion pipeline. """ images: Union[List[PIL.Image.Image], np.ndarray] guidance_images: Union[List[PIL.Image.Image], np.ndarray] class DiffuseHighSDXLPipeline(StableDiffusionXLPipeline): def __init__( self, vae: AutoencoderKL, text_encoder: CLIPTextModel, text_encoder_2: CLIPTextModelWithProjection, tokenizer: CLIPTokenizer, tokenizer_2: CLIPTokenizer, unet: UNet2DConditionModel, scheduler: KarrasDiffusionSchedulers, image_encoder: CLIPVisionModelWithProjection = None, feature_extractor: CLIPImageProcessor = None, force_zeros_for_empty_prompt: bool = True, add_watermarker: Optional[bool] = None, ): super().__init__( vae=vae, text_encoder=text_encoder, text_encoder_2=text_encoder_2, tokenizer=tokenizer, tokenizer_2=tokenizer_2, unet=unet, scheduler=scheduler, image_encoder=image_encoder, feature_extractor=feature_extractor, force_zeros_for_empty_prompt=force_zeros_for_empty_prompt, add_watermarker=add_watermarker ) def _encode_vae_image( self, image: torch.Tensor, normalize: bool = True, ): if normalize: image = image * 2 - 1 needs_upcasting = self.vae.dtype == torch.float16 and self.vae.config.force_upcast if needs_upcasting: self.upcast_vae() image = image.to(self.device) latents = self.vae.encode(image).latent_dist.mode() * self.vae.config.scaling_factor if needs_upcasting: self.vae.to(dtype=torch.float16) return latents.to(self.dtype) def _decode_vae_latent( self, latents: torch.Tensor, output_type: Optional[str] = 'pt', ): needs_upcasting = self.vae.dtype == torch.float16 and self.vae.config.force_upcast if needs_upcasting: self.upcast_vae() latents = latents.to(next(iter(self.vae.post_quant_conv.parameters())).dtype) latents = latents.to(self.device) image = self.vae.decode(latents / self.vae.config.scaling_factor, return_dict=False)[0] image = self.image_processor.postprocess(image, output_type=output_type) if needs_upcasting: self.vae.to(dtype=torch.float16) return image def edm_scheduler_step( self, model_output: torch.FloatTensor, timestep: Union[float, torch.FloatTensor], sample: torch.FloatTensor, s_churn: float = 0.0, s_tmin: float = 0.0, s_tmax: float = 0.0, s_noise: float = 1.0, LL_guidance: Optional[torch.FloatTensor] = None, generator: Optional[torch.Generator] = None, return_pred_original_sample: bool = False, ): assert isinstance(self.scheduler, EulerDiscreteScheduler) config = self.scheduler.config if self.scheduler.step_index is None: self.scheduler._init_step_index(timestep) step_index = self.scheduler.step_index sigma = self.scheduler.sigmas[step_index] gamma = min(s_churn / (len(self.scheduler.sigmas) - 1), 2**0.5 - 1) if s_tmin <= sigma <= s_tmax else 0.0 noise = randn_tensor( model_output.shape, dtype=model_output.dtype, device=model_output.device, generator=generator ) eps = noise * s_noise sigma_hat = sigma * (gamma + 1) if gamma > 0: sample = sample + eps * (sigma_hat**2 - sigma**2) ** 0.5 # 1. compute predicted original sample (x_0) from sigma-scaled predicted noise if config.prediction_type == "original_sample" or config.prediction_type == "sample": pred_original_sample = model_output elif config.prediction_type == "epsilon": pred_original_sample = sample - sigma_hat * model_output elif config.prediction_type == "v_prediction": # denoised = model_output * c_out + input * c_skip pred_original_sample = model_output * (-sigma / (sigma**2 + 1) ** 0.5) + (sample / (sigma**2 + 1)) else: raise ValueError( f"prediction_type given as {config.prediction_type} must be one of `epsilon`, or `v_prediction`" ) # 2. If gudiance LL component is given, perform structural guidance if LL_guidance is not None: pred_original_image = self._decode_vae_latent(pred_original_sample, output_type='pt') _, HH = self.DWT(pred_original_image) coeffs = (LL_guidance, HH) pred_original_image = self.iDWT(coeffs) pred_original_sample = self._encode_vae_image(pred_original_image) # 3. Convert to an ODE derivative derivative = (sample - pred_original_sample) / sigma_hat dt = self.scheduler.sigmas[self.scheduler.step_index + 1] - sigma_hat prev_sample = sample + derivative * dt self.scheduler._step_index += 1 if return_pred_original_sample: return (prev_sample, pred_original_sample) return (prev_sample, ) @torch.no_grad() def __call__( self, prompt: Union[str, List[str]] = None, prompt_2: Optional[Union[str, List[str]]] = None, num_inference_steps: int = 50, timesteps: List[int] = None, denoising_end: Optional[float] = None, guidance_scale: float = 5, negative_prompt: Optional[Union[str, List[str]]] = None, negative_prompt_2: Optional[Union[str, List[str]]] = None, num_images_per_prompt: Optional[int] = 1, eta: float = 0.0, generator: Optional[Union[torch.Generator, List[torch.Generator]]] = None, latents: Optional[torch.FloatTensor] = None, prompt_embeds: Optional[torch.FloatTensor] = None, negative_prompt_embeds: Optional[torch.FloatTensor] = None, pooled_prompt_embeds: Optional[torch.FloatTensor] = None, negative_pooled_prompt_embeds: Optional[torch.FloatTensor] = None, ip_adapter_image: Optional[PipelineImageInput] = None, output_type: Optional[str] = "pil", return_dict: bool = True, cross_attention_kwargs: Optional[Dict[str, Any]] = None, guidance_rescale: float = 0.0, original_size: Optional[Tuple[int, int]] = None, crops_coords_top_left: Tuple[int, int] = (0, 0), target_size: Optional[Tuple[int, int]] = None, negative_original_size: Optional[Tuple[int, int]] = None, negative_crops_coords_top_left: Tuple[int, int] = (0, 0), negative_target_size: Optional[Tuple[int, int]] = None, clip_skip: Optional[int] = None, callback: Optional[Callable[[int, int, torch.FloatTensor], None]] = None, callback_steps: Optional[int] = 1, callback_on_step_end: Optional[Callable[[int, int, Dict], None]] = None, callback_on_step_end_tensor_inputs: List[str] = ["latents"], ### DiffuseHigh parameters ### target_height: Union[int, List[int]] = [2048, 3072, 4096], target_width: Union[int, List[int]] = [2048, 3072, 4096], guidance_image: Optional[Union[torch.FloatTensor, PIL.Image.Image, np.ndarray]] = None, noising_steps: int = 15, diffusehigh_guidance_scale: float = 10.0, # >>> DWT parameters enable_dwt: bool = True, dwt_level: Optional[int] = 1, dwt_wave: Optional[str] = "db4", dwt_mode: Optional[str] = "symmetric", dwt_steps: Optional[int] = 5, # >>> Sharpening parameters enable_sharpening: bool = True, sharpening_kernel_size: int = 3, sharpening_sigma: Optional[Union[Tuple[float, float], float]] = (0.1, 2.0), sharpening_alpha: float = 1.0, **kwargs, ): r""" Function invoked when calling the pipeline for generation. Args: prompt (`str` or `List[str]`, *optional*): The prompt or prompts to guide the image generation. If not defined, one has to pass `prompt_embeds`. instead. prompt_2 (`str` or `List[str]`, *optional*): The prompt or prompts to be sent to the `tokenizer_2` and `text_encoder_2`. If not defined, `prompt` is used in both text-encoders height (`int`, *optional*, defaults to self.unet.config.sample_size * self.vae_scale_factor): The height in pixels of the generated image. This is set to 1024 by default for the best results. Anything below 512 pixels won't work well for [stabilityai/stable-diffusion-xl-base-1.0](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0) and checkpoints that are not specifically fine-tuned on low resolutions. width (`int`, *optional*, defaults to self.unet.config.sample_size * self.vae_scale_factor): The width in pixels of the generated image. This is set to 1024 by default for the best results. Anything below 512 pixels won't work well for [stabilityai/stable-diffusion-xl-base-1.0](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0) and checkpoints that are not specifically fine-tuned on low resolutions. num_inference_steps (`int`, *optional*, defaults to 50): The number of denoising steps. More denoising steps usually lead to a higher quality image at the expense of slower inference. timesteps (`List[int]`, *optional*): Custom timesteps to use for the denoising process with schedulers which support a `timesteps` argument in their `set_timesteps` method. If not defined, the default behavior when `num_inference_steps` is passed will be used. Must be in descending order. denoising_end (`float`, *optional*): When specified, determines the fraction (between 0.0 and 1.0) of the total denoising process to be completed before it is intentionally prematurely terminated. As a result, the returned sample will still retain a substantial amount of noise as determined by the discrete timesteps selected by the scheduler. The denoising_end parameter should ideally be utilized when this pipeline forms a part of a "Mixture of Denoisers" multi-pipeline setup, as elaborated in [**Refining the Image Output**](https://huggingface.co/docs/diffusers/api/pipelines/stable_diffusion/stable_diffusion_xl#refining-the-image-output) guidance_scale (`float`, *optional*, defaults to 5.0): Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598). `guidance_scale` is defined as `w` of equation 2. of [Imagen Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale > 1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`, usually at the expense of lower image quality. negative_prompt (`str` or `List[str]`, *optional*): The prompt or prompts not to guide the image generation. If not defined, one has to pass `negative_prompt_embeds` instead. Ignored when not using guidance (i.e., ignored if `guidance_scale` is less than `1`). negative_prompt_2 (`str` or `List[str]`, *optional*): The prompt or prompts not to guide the image generation to be sent to `tokenizer_2` and `text_encoder_2`. If not defined, `negative_prompt` is used in both text-encoders num_images_per_prompt (`int`, *optional*, defaults to 1): The number of images to generate per prompt. eta (`float`, *optional*, defaults to 0.0): Corresponds to parameter eta (η) in the DDIM paper: https://arxiv.org/abs/2010.02502. Only applies to [`schedulers.DDIMScheduler`], will be ignored for others. generator (`torch.Generator` or `List[torch.Generator]`, *optional*): One or a list of [torch generator(s)](https://pytorch.org/docs/stable/generated/torch.Generator.html) to make generation deterministic. latents (`torch.FloatTensor`, *optional*): Pre-generated noisy latents, sampled from a Gaussian distribution, to be used as inputs for image generation. Can be used to tweak the same generation with different prompts. If not provided, a latents tensor will ge generated by sampling using the supplied random `generator`. prompt_embeds (`torch.FloatTensor`, *optional*): Pre-generated text embeddings. Can be used to easily tweak text inputs, *e.g.* prompt weighting. If not provided, text embeddings will be generated from `prompt` input argument. negative_prompt_embeds (`torch.FloatTensor`, *optional*): Pre-generated negative text embeddings. Can be used to easily tweak text inputs, *e.g.* prompt weighting. If not provided, negative_prompt_embeds will be generated from `negative_prompt` input argument. pooled_prompt_embeds (`torch.FloatTensor`, *optional*): Pre-generated pooled text embeddings. Can be used to easily tweak text inputs, *e.g.* prompt weighting. If not provided, pooled text embeddings will be generated from `prompt` input argument. negative_pooled_prompt_embeds (`torch.FloatTensor`, *optional*): Pre-generated negative pooled text embeddings. Can be used to easily tweak text inputs, *e.g.* prompt weighting. If not provided, pooled negative_prompt_embeds will be generated from `negative_prompt` input argument. ip_adapter_image: (`PipelineImageInput`, *optional*): Optional image input to work with IP Adapters. output_type (`str`, *optional*, defaults to `"pil"`): The output format of the generate image. Choose between [PIL](https://pillow.readthedocs.io/en/stable/): `PIL.Image.Image` or `np.array`. return_dict (`bool`, *optional*, defaults to `True`): Whether or not to return a [`~pipelines.stable_diffusion_xl.StableDiffusionXLPipelineOutput`] instead of a plain tuple. cross_attention_kwargs (`dict`, *optional*): A kwargs dictionary that if specified is passed along to the `AttentionProcessor` as defined under `self.processor` in [diffusers.models.attention_processor](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention_processor.py). guidance_rescale (`float`, *optional*, defaults to 0.0): Guidance rescale factor proposed by [Common Diffusion Noise Schedules and Sample Steps are Flawed](https://arxiv.org/pdf/2305.08891.pdf) `guidance_scale` is defined as `φ` in equation 16. of [Common Diffusion Noise Schedules and Sample Steps are Flawed](https://arxiv.org/pdf/2305.08891.pdf). Guidance rescale factor should fix overexposure when using zero terminal SNR. original_size (`Tuple[int]`, *optional*, defaults to (1024, 1024)): If `original_size` is not the same as `target_size` the image will appear to be down- or upsampled. `original_size` defaults to `(height, width)` if not specified. Part of SDXL's micro-conditioning as explained in section 2.2 of [https://huggingface.co/papers/2307.01952](https://huggingface.co/papers/2307.01952). crops_coords_top_left (`Tuple[int]`, *optional*, defaults to (0, 0)): `crops_coords_top_left` can be used to generate an image that appears to be "cropped" from the position `crops_coords_top_left` downwards. Favorable, well-centered images are usually achieved by setting `crops_coords_top_left` to (0, 0). Part of SDXL's micro-conditioning as explained in section 2.2 of [https://huggingface.co/papers/2307.01952](https://huggingface.co/papers/2307.01952). target_size (`Tuple[int]`, *optional*, defaults to (1024, 1024)): For most cases, `target_size` should be set to the desired height and width of the generated image. If not specified it will default to `(height, width)`. Part of SDXL's micro-conditioning as explained in section 2.2 of [https://huggingface.co/papers/2307.01952](https://huggingface.co/papers/2307.01952). negative_original_size (`Tuple[int]`, *optional*, defaults to (1024, 1024)): To negatively condition the generation process based on a specific image resolution. Part of SDXL's micro-conditioning as explained in section 2.2 of [https://huggingface.co/papers/2307.01952](https://huggingface.co/papers/2307.01952). For more information, refer to this issue thread: https://github.com/huggingface/diffusers/issues/4208. negative_crops_coords_top_left (`Tuple[int]`, *optional*, defaults to (0, 0)): To negatively condition the generation process based on a specific crop coordinates. Part of SDXL's micro-conditioning as explained in section 2.2 of [https://huggingface.co/papers/2307.01952](https://huggingface.co/papers/2307.01952). For more information, refer to this issue thread: https://github.com/huggingface/diffusers/issues/4208. negative_target_size (`Tuple[int]`, *optional*, defaults to (1024, 1024)): To negatively condition the generation process based on a target image resolution. It should be as same as the `target_size` for most cases. Part of SDXL's micro-conditioning as explained in section 2.2 of [https://huggingface.co/papers/2307.01952](https://huggingface.co/papers/2307.01952). For more information, refer to this issue thread: https://github.com/huggingface/diffusers/issues/4208. callback_on_step_end (`Callable`, *optional*): A function that calls at the end of each denoising steps during the inference. The function is called with the following arguments: `callback_on_step_end(self: DiffusionPipeline, step: int, timestep: int, callback_kwargs: Dict)`. `callback_kwargs` will include a list of all tensors as specified by `callback_on_step_end_tensor_inputs`. callback_on_step_end_tensor_inputs (`List`, *optional*): The list of tensor inputs for the `callback_on_step_end` function. The tensors specified in the list will be passed as `callback_kwargs` argument. You will only be able to include variables listed in the `._callback_tensor_inputs` attribute of your pipeline class. target_height ('List[int]' or int): The height of the image being generated. If list is given, the pipeline generates corresponding intermediate resolution images in a progressive manner. target_width ('List[int]' or int): The width of the image being generated. If list is given, the pipeline generates corresponding intermediate resolution images in a progressive manner. Examples: Returns: [`DiffuseHighSDXLPipelineOutput`] or `tuple`: [`DiffuseHighSDXLPipelineOutput`] if `return_dict` is True, otherwise a `tuple`. When returning a tuple, the first element is a list with the generated images. """ # 0. Default height and width to unet height = self.default_sample_size * self.vae_scale_factor width = self.default_sample_size * self.vae_scale_factor original_size = original_size or (height, width) target_size = target_size or (height, width) # 1. Check inputs. Raise error if not correct self.check_inputs( prompt, prompt_2, height, width, callback_steps, negative_prompt, negative_prompt_2, prompt_embeds, negative_prompt_embeds, pooled_prompt_embeds, negative_pooled_prompt_embeds, callback_on_step_end_tensor_inputs, ) self._guidance_scale = guidance_scale self._guidance_rescale = guidance_rescale self._clip_skip = clip_skip self._cross_attention_kwargs = cross_attention_kwargs self._denoising_end = denoising_end # 2. Define call parameters if prompt is not None and isinstance(prompt, str): batch_size = 1 elif prompt is not None and isinstance(prompt, list): batch_size = len(prompt) else: batch_size = prompt_embeds.shape[0] device = self._execution_device # 3. Encode input prompt lora_scale = ( self.cross_attention_kwargs.get("scale", None) if self.cross_attention_kwargs is not None else None ) ( prompt_embeds, negative_prompt_embeds, pooled_prompt_embeds, negative_pooled_prompt_embeds, ) = self.encode_prompt( prompt=prompt, prompt_2=prompt_2, device=device, num_images_per_prompt=num_images_per_prompt, do_classifier_free_guidance=self.do_classifier_free_guidance, negative_prompt=negative_prompt, negative_prompt_2=negative_prompt_2, prompt_embeds=prompt_embeds, negative_prompt_embeds=negative_prompt_embeds, pooled_prompt_embeds=pooled_prompt_embeds, negative_pooled_prompt_embeds=negative_pooled_prompt_embeds, lora_scale=lora_scale, clip_skip=self.clip_skip, ) # 4. Prepare timesteps timesteps, num_inference_steps = retrieve_timesteps(self.scheduler, num_inference_steps, device, timesteps) # 5. Prepare latent variables num_channels_latents = self.unet.config.in_channels latents = self.prepare_latents( batch_size * num_images_per_prompt, num_channels_latents, height, width, prompt_embeds.dtype, device, generator, latents, ) # 6. Prepare extra step kwargs. TODO: Logic should ideally just be moved out of the pipeline extra_step_kwargs = self.prepare_extra_step_kwargs(generator, eta) # 7. Prepare added time ids & embeddings add_text_embeds = pooled_prompt_embeds if self.text_encoder_2 is None: text_encoder_projection_dim = int(pooled_prompt_embeds.shape[-1]) else: text_encoder_projection_dim = self.text_encoder_2.config.projection_dim add_time_ids = self._get_add_time_ids( original_size, crops_coords_top_left, target_size, dtype=prompt_embeds.dtype, text_encoder_projection_dim=text_encoder_projection_dim, ) if negative_original_size is not None and negative_target_size is not None: negative_add_time_ids = self._get_add_time_ids( negative_original_size, negative_crops_coords_top_left, negative_target_size, dtype=prompt_embeds.dtype, text_encoder_projection_dim=text_encoder_projection_dim, ) else: negative_add_time_ids = add_time_ids if self.do_classifier_free_guidance: prompt_embeds = torch.cat([negative_prompt_embeds, prompt_embeds], dim=0) add_text_embeds = torch.cat([negative_pooled_prompt_embeds, add_text_embeds], dim=0) add_time_ids = torch.cat([negative_add_time_ids, add_time_ids], dim=0) prompt_embeds = prompt_embeds.to(device) add_text_embeds = add_text_embeds.to(device) add_time_ids = add_time_ids.to(device).repeat(batch_size * num_images_per_prompt, 1) if ip_adapter_image is not None: image_embeds, negative_image_embeds = self.encode_image(ip_adapter_image, device, num_images_per_prompt) if self.do_classifier_free_guidance: image_embeds = torch.cat([negative_image_embeds, image_embeds]) image_embeds = image_embeds.to(device) # 8. Denoising loop num_warmup_steps = max(len(timesteps) - num_inference_steps * self.scheduler.order, 0) # 8.1 Apply denoising_end if ( self.denoising_end is not None and isinstance(self.denoising_end, float) and self.denoising_end > 0 and self.denoising_end < 1 ): discrete_timestep_cutoff = int( round( self.scheduler.config.num_train_timesteps - (self.denoising_end * self.scheduler.config.num_train_timesteps) ) ) num_inference_steps = len(list(filter(lambda ts: ts >= discrete_timestep_cutoff, timesteps))) timesteps = timesteps[:num_inference_steps] # 9. Optionally get Guidance Scale Embedding timestep_cond = None if self.unet.config.time_cond_proj_dim is not None: guidance_scale_tensor = torch.tensor(self.guidance_scale - 1).repeat(batch_size * num_images_per_prompt) timestep_cond = self.get_guidance_scale_embedding( guidance_scale_tensor, embedding_dim=self.unet.config.time_cond_proj_dim ).to(device=device, dtype=latents.dtype) # 10. Obtain clean image for structral guidance (can be given by user or generated) if guidance_image is None: self._num_timesteps = len(timesteps) with self.progress_bar(total=num_inference_steps) as progress_bar: for i, t in enumerate(timesteps): # expand the latents if we are doing classifier free guidance latent_model_input = torch.cat([latents] * 2) if self.do_classifier_free_guidance else latents latent_model_input = self.scheduler.scale_model_input(latent_model_input, t) # predict the noise residual added_cond_kwargs = {"text_embeds": add_text_embeds, "time_ids": add_time_ids} if ip_adapter_image is not None: added_cond_kwargs["image_embeds"] = image_embeds noise_pred = self.unet( latent_model_input, t, encoder_hidden_states=prompt_embeds, timestep_cond=timestep_cond, cross_attention_kwargs=self.cross_attention_kwargs, added_cond_kwargs=added_cond_kwargs, return_dict=False, )[0] # perform guidance if self.do_classifier_free_guidance: noise_pred_uncond, noise_pred_text = noise_pred.chunk(2) noise_pred = noise_pred_uncond + self.guidance_scale * (noise_pred_text - noise_pred_uncond) if self.do_classifier_free_guidance and self.guidance_rescale > 0.0: # Based on 3.4. in https://arxiv.org/pdf/2305.08891.pdf noise_pred = rescale_noise_cfg(noise_pred, noise_pred_text, guidance_rescale=self.guidance_rescale) # compute the previous noisy sample x_t -> x_t-1 latents = self.scheduler.step(noise_pred, t, latents, **extra_step_kwargs, return_dict=False)[0] if callback_on_step_end is not None: callback_kwargs = {} for k in callback_on_step_end_tensor_inputs: callback_kwargs[k] = locals()[k] callback_outputs = callback_on_step_end(self, i, t, callback_kwargs) latents = callback_outputs.pop("latents", latents) prompt_embeds = callback_outputs.pop("prompt_embeds", prompt_embeds) negative_prompt_embeds = callback_outputs.pop("negative_prompt_embeds", negative_prompt_embeds) add_text_embeds = callback_outputs.pop("add_text_embeds", add_text_embeds) negative_pooled_prompt_embeds = callback_outputs.pop( "negative_pooled_prompt_embeds", negative_pooled_prompt_embeds ) add_time_ids = callback_outputs.pop("add_time_ids", add_time_ids) negative_add_time_ids = callback_outputs.pop("negative_add_time_ids", negative_add_time_ids) # call the callback, if provided if i == len(timesteps) - 1 or ((i + 1) > num_warmup_steps and (i + 1) % self.scheduler.order == 0): progress_bar.update() if callback is not None and i % callback_steps == 0: step_idx = i // getattr(self.scheduler, "order", 1) callback(step_idx, t, latents) image = self._decode_vae_latent(latents, output_type='pt') else: image = self.image_processor.preprocess(guidance_image, height, width) if self.image_processor.config.do_normalize: image = (image + 1.) * 0.5 image = image.to(self.device) original_guidance_image = image # |-------------------------------- DiffuseHigh process --------------------------------| # DWT & inverse DWT works on torch.float32 if enable_dwt: self.DWT = DWTForward(J=dwt_level, wave=dwt_wave, mode=dwt_mode).to(self.device) self.iDWT = DWTInverse(wave=dwt_wave, mode=dwt_mode).to(self.device) # 11. Prepare progressive DiffuseHigh pipeline self.scheduler.set_timesteps(num_inference_steps) diffusehigh_timesteps = self.scheduler.timesteps[-noising_steps:] self.enable_vae_tiling() # Vae tiling mode in order to prevent OOM issues if isinstance(target_width, int): target_width = [target_width] if isinstance(target_height, int): target_height = [target_height] assert len(target_width) == len(target_height) #12. Progressive DiffuseHigh Pipeline for h, w in zip(target_height, target_width): # interpolate the image to the desired resolution guidance_image = F.interpolate(image, (h, w), mode="bicubic", align_corners=False) # apply sharpening operation to the image if enable_sharpening: guidance_image = gaussian_blur_image_sharpening( guidance_image, kernel_size=sharpening_kernel_size, sigma=sharpening_sigma, alpha=sharpening_alpha, ) # extract low-frequency component (structural guidance) from the guidance image if enable_dwt: LL, _ = self.DWT(guidance_image) # obtain latent of the interpolated image and noise it latents = self._encode_vae_image(guidance_image) noise = randn_tensor(latents.shape, generator, device=latents.device, dtype=latents.dtype) latents = self.scheduler.add_noise(latents, noise, diffusehigh_timesteps[None, 0]) for i, t in tqdm(enumerate(diffusehigh_timesteps), total=diffusehigh_timesteps.shape[0]): # expand the latents if we are doing classifier free guidance latent_model_input = torch.cat([latents] * 2) if self.do_classifier_free_guidance else latents latent_model_input = self.scheduler.scale_model_input(latent_model_input, t) # predict the noise residual added_cond_kwargs = {"text_embeds": add_text_embeds, "time_ids": add_time_ids} noise_pred = self.unet( latent_model_input, t, encoder_hidden_states=prompt_embeds, timestep_cond=timestep_cond, cross_attention_kwargs=self.cross_attention_kwargs, # None added_cond_kwargs=added_cond_kwargs, # None return_dict=False, )[0] # perform guidance if self.do_classifier_free_guidance: noise_pred_uncond, noise_pred_text = noise_pred.chunk(2) noise_pred = noise_pred_uncond + diffusehigh_guidance_scale * (noise_pred_text - noise_pred_uncond) # EDM sampler step latents = self.edm_scheduler_step( noise_pred, t, latents, **extra_step_kwargs, LL_guidance=LL if (enable_dwt and i < dwt_steps) else None, )[0] image = self._decode_vae_latent(latents) if isinstance(self.scheduler, EulerDiscreteScheduler): self.scheduler._step_index = None # Offload all models self.maybe_free_model_hooks() if output_type != 'pt': image = self.image_processor.postprocess(image * 2 - 1, output_type=output_type) guidance_image = self.image_processor.postprocess(original_guidance_image * 2 -1 , output_type=output_type) if not return_dict: return (image, guidance_image) return DiffuseHighSDXLPipelineOutput(images=image, guidance_image=guidance_image) def set_seeds(seed): os.environ["PYTHONHASHSEED"] = str(seed) np.random.seed(seed) torch.manual_seed(seed) torch.cuda.manual_seed(seed) torch.backends.cudnn.deterministic = True torch.backends.cudnn.benchmark = True # DEBUGGING if __name__ == "__main__": set_seeds(23) model = DiffuseHighSDXLPipeline.from_pretrained( "stabilityai/stable-diffusion-xl-base-1.0", torch_dtype=torch.float16, #scheduler=scheduler ).to("cuda") prompt = "Cinematic photo of delicious chocolate icecream." negative_prompt = "blurry, ugly, duplicate, poorly drawn, deformed, mosaic" image = model( prompt, negative_prompt=negative_prompt, target_height=[2048, 3072, 4096], target_width=[2048, 3072, 4096], enable_dwt=True, dwt_steps=5, enable_sharpening=True, sharpness_factor=1.0, ).images[0] image.save("sample.png")