# Copyright 2024 Stability AI and The HuggingFace Team. All rights reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. from typing import Any, Callable, Dict, List, Optional, Union import PIL.Image from diffusers.image_processor import PipelineImageInput, VaeImageProcessor from diffusers.loaders import FromSingleFileMixin, SD3LoraLoaderMixin from diffusers.models.autoencoders import AutoencoderKL from diffusers.models.transformers import SD3Transformer2DModel from diffusers.pipelines.pipeline_utils import DiffusionPipeline from diffusers.pipelines.stable_diffusion_3.pipeline_output import ( StableDiffusion3PipelineOutput, ) from diffusers.pipelines.stable_diffusion_3.pipeline_stable_diffusion_3_img2img import ( retrieve_latents, retrieve_timesteps, ) from diffusers.schedulers import FlowMatchEulerDiscreteScheduler from diffusers.utils import is_torch_xla_available, logging, replace_example_docstring from diffusers.utils import deprecate, logging from diffusers.utils.torch_utils import randn_tensor import torch from transformers import ( CLIPTextModelWithProjection, CLIPTokenizer, T5EncoderModel, T5TokenizerFast, ) if is_torch_xla_available(): import torch_xla.core.xla_model as xm XLA_AVAILABLE = True else: XLA_AVAILABLE = False logger = logging.get_logger(__name__) # pylint: disable=invalid-name EXAMPLE_DOC_STRING = """ Examples: ```py >>> import torch >>> from diffusers import StableDiffusion3Pipeline >>> pipe = StableDiffusion3Pipeline.from_pretrained( ... "stabilityai/stable-diffusion-3-medium-diffusers", torch_dtype=torch.float16 ... ) >>> pipe.to("cuda") >>> prompt = "A cat holding a sign that says hello world" >>> image = pipe(prompt).images[0] >>> image.save("sd3.png") ``` """ class StableDiffusion3InstructPix2PixPipeline(DiffusionPipeline, SD3LoraLoaderMixin, FromSingleFileMixin): r""" Args: transformer ([`SD3Transformer2DModel`]): Conditional Transformer (MMDiT) architecture to denoise the encoded image latents. scheduler ([`FlowMatchEulerDiscreteScheduler`]): A scheduler to be used in combination with `transformer` to denoise the encoded image latents. vae ([`AutoencoderKL`]): Variational Auto-Encoder (VAE) Model to encode and decode images to and from latent representations. text_encoder ([`CLIPTextModelWithProjection`]): [CLIP](https://huggingface.co/docs/transformers/model_doc/clip#transformers.CLIPTextModelWithProjection), specifically the [clip-vit-large-patch14](https://huggingface.co/openai/clip-vit-large-patch14) variant, with an additional added projection layer that is initialized with a diagonal matrix with the `hidden_size` as its dimension. text_encoder_2 ([`CLIPTextModelWithProjection`]): [CLIP](https://huggingface.co/docs/transformers/model_doc/clip#transformers.CLIPTextModelWithProjection), specifically the [laion/CLIP-ViT-bigG-14-laion2B-39B-b160k](https://huggingface.co/laion/CLIP-ViT-bigG-14-laion2B-39B-b160k) variant. text_encoder_3 ([`T5EncoderModel`]): Frozen text-encoder. Stable Diffusion 3 uses [T5](https://huggingface.co/docs/transformers/model_doc/t5#transformers.T5EncoderModel), specifically the [t5-v1_1-xxl](https://huggingface.co/google/t5-v1_1-xxl) variant. tokenizer (`CLIPTokenizer`): Tokenizer of class [CLIPTokenizer](https://huggingface.co/docs/transformers/v4.21.0/en/model_doc/clip#transformers.CLIPTokenizer). tokenizer_2 (`CLIPTokenizer`): Second Tokenizer of class [CLIPTokenizer](https://huggingface.co/docs/transformers/v4.21.0/en/model_doc/clip#transformers.CLIPTokenizer). tokenizer_3 (`T5TokenizerFast`): Tokenizer of class [T5Tokenizer](https://huggingface.co/docs/transformers/model_doc/t5#transformers.T5Tokenizer). """ model_cpu_offload_seq = "text_encoder->text_encoder_2->text_encoder_3->transformer->vae" _optional_components = [] _callback_tensor_inputs = ["latents", "prompt_embeds", "negative_prompt_embeds", "negative_pooled_prompt_embeds"] def __init__( self, transformer: SD3Transformer2DModel, scheduler: FlowMatchEulerDiscreteScheduler, vae: AutoencoderKL, text_encoder: CLIPTextModelWithProjection, tokenizer: CLIPTokenizer, text_encoder_2: CLIPTextModelWithProjection, tokenizer_2: CLIPTokenizer, text_encoder_3: T5EncoderModel, tokenizer_3: T5TokenizerFast, ): super().__init__() self.register_modules( vae=vae, text_encoder=text_encoder, text_encoder_2=text_encoder_2, text_encoder_3=text_encoder_3, tokenizer=tokenizer, tokenizer_2=tokenizer_2, tokenizer_3=tokenizer_3, transformer=transformer, scheduler=scheduler, ) self.vae_scale_factor = ( 2 ** (len(self.vae.config.block_out_channels) - 1) if hasattr(self, "vae") and self.vae is not None else 8 ) self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor) self.tokenizer_max_length = ( self.tokenizer.model_max_length if hasattr(self, "tokenizer") and self.tokenizer is not None else 77 ) self.default_sample_size = ( self.transformer.config.sample_size if hasattr(self, "transformer") and self.transformer is not None else 128 ) def _get_t5_prompt_embeds( self, prompt: Union[str, List[str]] = None, num_images_per_prompt: int = 1, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None, ): device = device or self._execution_device dtype = dtype or self.text_encoder.dtype prompt = [prompt] if isinstance(prompt, str) else prompt batch_size = len(prompt) if self.text_encoder_3 is None: return torch.zeros( (batch_size, self.tokenizer_max_length, self.transformer.config.joint_attention_dim), device=device, dtype=dtype, ) text_inputs = self.tokenizer_3( prompt, padding="max_length", max_length=self.tokenizer_max_length, truncation=True, add_special_tokens=True, return_tensors="pt", ) text_input_ids = text_inputs.input_ids untruncated_ids = self.tokenizer_3(prompt, padding="longest", return_tensors="pt").input_ids if untruncated_ids.shape[-1] >= text_input_ids.shape[-1] and not torch.equal(text_input_ids, untruncated_ids): removed_text = self.tokenizer_3.batch_decode(untruncated_ids[:, self.tokenizer_max_length - 1 : -1]) logger.warning( "The following part of your input was truncated because CLIP can only handle sequences up to" f" {self.tokenizer_max_length} tokens: {removed_text}" ) prompt_embeds = self.text_encoder_3(text_input_ids.to(device))[0] dtype = self.text_encoder_3.dtype prompt_embeds = prompt_embeds.to(dtype=dtype, device=device) _, seq_len, _ = prompt_embeds.shape # duplicate text embeddings and attention mask for each generation per prompt, using mps friendly method prompt_embeds = prompt_embeds.repeat(1, num_images_per_prompt, 1) prompt_embeds = prompt_embeds.view(batch_size * num_images_per_prompt, seq_len, -1) return prompt_embeds def _get_clip_prompt_embeds( self, prompt: Union[str, List[str]], num_images_per_prompt: int = 1, device: Optional[torch.device] = None, clip_skip: Optional[int] = None, clip_model_index: int = 0, ): device = device or self._execution_device clip_tokenizers = [self.tokenizer, self.tokenizer_2] clip_text_encoders = [self.text_encoder, self.text_encoder_2] tokenizer = clip_tokenizers[clip_model_index] text_encoder = clip_text_encoders[clip_model_index] prompt = [prompt] if isinstance(prompt, str) else prompt batch_size = len(prompt) text_inputs = tokenizer( prompt, padding="max_length", max_length=self.tokenizer_max_length, truncation=True, return_tensors="pt", ) text_input_ids = text_inputs.input_ids untruncated_ids = tokenizer(prompt, padding="longest", return_tensors="pt").input_ids if untruncated_ids.shape[-1] >= text_input_ids.shape[-1] and not torch.equal(text_input_ids, untruncated_ids): removed_text = tokenizer.batch_decode(untruncated_ids[:, self.tokenizer_max_length - 1 : -1]) logger.warning( "The following part of your input was truncated because CLIP can only handle sequences up to" f" {self.tokenizer_max_length} tokens: {removed_text}" ) prompt_embeds = text_encoder(text_input_ids.to(device), output_hidden_states=True) pooled_prompt_embeds = prompt_embeds[0] if clip_skip is None: prompt_embeds = prompt_embeds.hidden_states[-2] else: prompt_embeds = prompt_embeds.hidden_states[-(clip_skip + 2)] prompt_embeds = prompt_embeds.to(dtype=self.text_encoder.dtype, device=device) _, seq_len, _ = prompt_embeds.shape # duplicate text embeddings for each generation per prompt, using mps friendly method prompt_embeds = prompt_embeds.repeat(1, num_images_per_prompt, 1) prompt_embeds = prompt_embeds.view(batch_size * num_images_per_prompt, seq_len, -1) pooled_prompt_embeds = pooled_prompt_embeds.repeat(1, num_images_per_prompt, 1) pooled_prompt_embeds = pooled_prompt_embeds.view(batch_size * num_images_per_prompt, -1) return prompt_embeds, pooled_prompt_embeds def encode_prompt( self, prompt: Union[str, List[str]], prompt_2: Union[str, List[str]], prompt_3: Union[str, List[str]], device: Optional[torch.device] = None, num_images_per_prompt: int = 1, do_classifier_free_guidance: bool = True, negative_prompt: Optional[Union[str, List[str]]] = None, negative_prompt_2: Optional[Union[str, List[str]]] = None, negative_prompt_3: Optional[Union[str, List[str]]] = 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, clip_skip: Optional[int] = None, ): r""" Args: prompt (`str` or `List[str]`, *optional*): prompt to be encoded 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 all text-encoders prompt_3 (`str` or `List[str]`, *optional*): The prompt or prompts to be sent to the `tokenizer_3` and `text_encoder_3`. If not defined, `prompt` is used in all text-encoders device: (`torch.device`): torch device num_images_per_prompt (`int`): number of images that should be generated per prompt do_classifier_free_guidance (`bool`): whether to use classifier free guidance or not 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 all the text-encoders. negative_prompt_2 (`str` or `List[str]`, *optional*): The prompt or prompts not to guide the image generation to be sent to `tokenizer_3` and `text_encoder_3`. If not defined, `negative_prompt` is used in both text-encoders 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. clip_skip (`int`, *optional*): Number of layers to be skipped from CLIP while computing the prompt embeddings. A value of 1 means that the output of the pre-final layer will be used for computing the prompt embeddings. """ device = device or self._execution_device prompt = [prompt] if isinstance(prompt, str) else prompt if prompt is not None: batch_size = len(prompt) else: batch_size = prompt_embeds.shape[0] if prompt_embeds is None: prompt_2 = prompt_2 or prompt prompt_2 = [prompt_2] if isinstance(prompt_2, str) else prompt_2 prompt_3 = prompt_3 or prompt prompt_3 = [prompt_3] if isinstance(prompt_3, str) else prompt_3 prompt_embed, pooled_prompt_embed = self._get_clip_prompt_embeds( prompt=prompt, device=device, num_images_per_prompt=num_images_per_prompt, clip_skip=clip_skip, clip_model_index=0, ) prompt_2_embed, pooled_prompt_2_embed = self._get_clip_prompt_embeds( prompt=prompt_2, device=device, num_images_per_prompt=num_images_per_prompt, clip_skip=clip_skip, clip_model_index=1, ) clip_prompt_embeds = torch.cat([prompt_embed, prompt_2_embed], dim=-1) t5_prompt_embed = self._get_t5_prompt_embeds( prompt=prompt_3, num_images_per_prompt=num_images_per_prompt, device=device, ) clip_prompt_embeds = torch.nn.functional.pad( clip_prompt_embeds, (0, t5_prompt_embed.shape[-1] - clip_prompt_embeds.shape[-1]) ) prompt_embeds = torch.cat([clip_prompt_embeds, t5_prompt_embed], dim=-2) pooled_prompt_embeds = torch.cat([pooled_prompt_embed, pooled_prompt_2_embed], dim=-1) if do_classifier_free_guidance and negative_prompt_embeds is None: negative_prompt = negative_prompt or "" negative_prompt_2 = negative_prompt_2 or negative_prompt negative_prompt_3 = negative_prompt_3 or negative_prompt # normalize str to list negative_prompt = batch_size * [negative_prompt] if isinstance(negative_prompt, str) else negative_prompt negative_prompt_2 = ( batch_size * [negative_prompt_2] if isinstance(negative_prompt_2, str) else negative_prompt_2 ) negative_prompt_3 = ( batch_size * [negative_prompt_3] if isinstance(negative_prompt_3, str) else negative_prompt_3 ) if prompt is not None and type(prompt) is not type(negative_prompt): raise TypeError( f"`negative_prompt` should be the same type to `prompt`, but got {type(negative_prompt)} !=" f" {type(prompt)}." ) elif batch_size != len(negative_prompt): raise ValueError( f"`negative_prompt`: {negative_prompt} has batch size {len(negative_prompt)}, but `prompt`:" f" {prompt} has batch size {batch_size}. Please make sure that passed `negative_prompt` matches" " the batch size of `prompt`." ) negative_prompt_embed, negative_pooled_prompt_embed = self._get_clip_prompt_embeds( negative_prompt, device=device, num_images_per_prompt=num_images_per_prompt, clip_skip=None, clip_model_index=0, ) negative_prompt_2_embed, negative_pooled_prompt_2_embed = self._get_clip_prompt_embeds( negative_prompt_2, device=device, num_images_per_prompt=num_images_per_prompt, clip_skip=None, clip_model_index=1, ) negative_clip_prompt_embeds = torch.cat([negative_prompt_embed, negative_prompt_2_embed], dim=-1) t5_negative_prompt_embed = self._get_t5_prompt_embeds( prompt=negative_prompt_3, num_images_per_prompt=num_images_per_prompt, device=device ) negative_clip_prompt_embeds = torch.nn.functional.pad( negative_clip_prompt_embeds, (0, t5_negative_prompt_embed.shape[-1] - negative_clip_prompt_embeds.shape[-1]), ) negative_prompt_embeds = torch.cat([negative_clip_prompt_embeds, t5_negative_prompt_embed], dim=-2) negative_pooled_prompt_embeds = torch.cat( [negative_pooled_prompt_embed, negative_pooled_prompt_2_embed], dim=-1 ) return prompt_embeds, negative_prompt_embeds, pooled_prompt_embeds, negative_pooled_prompt_embeds def check_inputs( self, prompt, prompt_2, prompt_3, # height, # width, negative_prompt=None, negative_prompt_2=None, negative_prompt_3=None, prompt_embeds=None, negative_prompt_embeds=None, pooled_prompt_embeds=None, negative_pooled_prompt_embeds=None, callback_on_step_end_tensor_inputs=None, ): # if height % 8 != 0 or width % 8 != 0: # raise ValueError(f"`height` and `width` have to be divisible by 8 but are {height} and {width}.") if callback_on_step_end_tensor_inputs is not None and not all( k in self._callback_tensor_inputs for k in callback_on_step_end_tensor_inputs ): raise ValueError( f"`callback_on_step_end_tensor_inputs` has to be in {self._callback_tensor_inputs}, but found {[k for k in callback_on_step_end_tensor_inputs if k not in self._callback_tensor_inputs]}" ) if prompt is not None and prompt_embeds is not None: raise ValueError( f"Cannot forward both `prompt`: {prompt} and `prompt_embeds`: {prompt_embeds}. Please make sure to" " only forward one of the two." ) elif prompt_2 is not None and prompt_embeds is not None: raise ValueError( f"Cannot forward both `prompt_2`: {prompt_2} and `prompt_embeds`: {prompt_embeds}. Please make sure to" " only forward one of the two." ) elif prompt_3 is not None and prompt_embeds is not None: raise ValueError( f"Cannot forward both `prompt_3`: {prompt_2} and `prompt_embeds`: {prompt_embeds}. Please make sure to" " only forward one of the two." ) elif prompt is None and prompt_embeds is None: raise ValueError( "Provide either `prompt` or `prompt_embeds`. Cannot leave both `prompt` and `prompt_embeds` undefined." ) elif prompt is not None and (not isinstance(prompt, str) and not isinstance(prompt, list)): raise ValueError(f"`prompt` has to be of type `str` or `list` but is {type(prompt)}") elif prompt_2 is not None and (not isinstance(prompt_2, str) and not isinstance(prompt_2, list)): raise ValueError(f"`prompt_2` has to be of type `str` or `list` but is {type(prompt_2)}") elif prompt_3 is not None and (not isinstance(prompt_3, str) and not isinstance(prompt_3, list)): raise ValueError(f"`prompt_3` has to be of type `str` or `list` but is {type(prompt_3)}") if negative_prompt is not None and negative_prompt_embeds is not None: raise ValueError( f"Cannot forward both `negative_prompt`: {negative_prompt} and `negative_prompt_embeds`:" f" {negative_prompt_embeds}. Please make sure to only forward one of the two." ) elif negative_prompt_2 is not None and negative_prompt_embeds is not None: raise ValueError( f"Cannot forward both `negative_prompt_2`: {negative_prompt_2} and `negative_prompt_embeds`:" f" {negative_prompt_embeds}. Please make sure to only forward one of the two." ) elif negative_prompt_3 is not None and negative_prompt_embeds is not None: raise ValueError( f"Cannot forward both `negative_prompt_3`: {negative_prompt_3} and `negative_prompt_embeds`:" f" {negative_prompt_embeds}. Please make sure to only forward one of the two." ) if prompt_embeds is not None and negative_prompt_embeds is not None: if prompt_embeds.shape != negative_prompt_embeds.shape: raise ValueError( "`prompt_embeds` and `negative_prompt_embeds` must have the same shape when passed directly, but" f" got: `prompt_embeds` {prompt_embeds.shape} != `negative_prompt_embeds`" f" {negative_prompt_embeds.shape}." ) if prompt_embeds is not None and pooled_prompt_embeds is None: raise ValueError( "If `prompt_embeds` are provided, `pooled_prompt_embeds` also have to be passed. Make sure to generate `pooled_prompt_embeds` from the same text encoder that was used to generate `prompt_embeds`." ) if negative_prompt_embeds is not None and negative_pooled_prompt_embeds is None: raise ValueError( "If `negative_prompt_embeds` are provided, `negative_pooled_prompt_embeds` also have to be passed. Make sure to generate `negative_pooled_prompt_embeds` from the same text encoder that was used to generate `negative_prompt_embeds`." ) def get_timesteps(self, num_inference_steps, strength, device): # get the original timestep using init_timestep init_timestep = min(num_inference_steps * strength, num_inference_steps) t_start = int(max(num_inference_steps - init_timestep, 0)) timesteps = self.scheduler.timesteps[t_start * self.scheduler.order :] if hasattr(self.scheduler, "set_begin_index"): self.scheduler.set_begin_index(t_start * self.scheduler.order) return timesteps, num_inference_steps - t_start def prepare_latents(self, image, timestep, batch_size, num_images_per_prompt, dtype, device, generator=None): if not isinstance(image, (torch.Tensor, PIL.Image.Image, list)): raise ValueError( f"`image` has to be of type `torch.Tensor`, `PIL.Image.Image` or list but is {type(image)}" ) image = image.to(device=device, dtype=dtype) batch_size = batch_size * num_images_per_prompt if image.shape[1] == self.vae.config.latent_channels: init_latents = image else: if isinstance(generator, list) and len(generator) != batch_size: raise ValueError( f"You have passed a list of generators of length {len(generator)}, but requested an effective batch" f" size of {batch_size}. Make sure the batch size matches the length of the generators." ) elif isinstance(generator, list): init_latents = [ retrieve_latents(self.vae.encode(image[i : i + 1]), generator=generator[i]) for i in range(batch_size) ] init_latents = torch.cat(init_latents, dim=0) else: init_latents = retrieve_latents(self.vae.encode(image), generator=generator) init_latents = (init_latents - self.vae.config.shift_factor) * self.vae.config.scaling_factor if batch_size > init_latents.shape[0] and batch_size % init_latents.shape[0] == 0: # expand init_latents for batch_size additional_image_per_prompt = batch_size // init_latents.shape[0] init_latents = torch.cat([init_latents] * additional_image_per_prompt, dim=0) elif batch_size > init_latents.shape[0] and batch_size % init_latents.shape[0] != 0: raise ValueError( f"Cannot duplicate `image` of batch size {init_latents.shape[0]} to {batch_size} text prompts." ) else: init_latents = torch.cat([init_latents], dim=0) shape = init_latents.shape noise = randn_tensor(shape, generator=generator, device=device, dtype=dtype) # get latents init_latents = self.scheduler.scale_noise(init_latents, timestep, noise) latents = init_latents.to(device=device, dtype=dtype) return latents def prepare_image_latents( self, image, batch_size, num_images_per_prompt, dtype, device, do_classifier_free_guidance, generator=None ): if not isinstance(image, (torch.Tensor, PIL.Image.Image, list)): raise ValueError( f"`image` has to be of type `torch.Tensor`, `PIL.Image.Image` or list but is {type(image)}" ) image = image.to(device=device, dtype=dtype) batch_size = batch_size * num_images_per_prompt if image.shape[1] == self.vae.config.latent_channels: image_latents = image else: image_latents = retrieve_latents(self.vae.encode(image), sample_mode="argmax") # ? normalize image latents # image_latents = (image_latents - self.vae.config.shift_factor) * self.vae.config.scaling_factor if batch_size > image_latents.shape[0] and batch_size % image_latents.shape[0] == 0: # expand image_latents for batch_size deprecation_message = ( f"You have passed {batch_size} text prompts (`prompt`), but only {image_latents.shape[0]} initial" " images (`image`). Initial images are now duplicating to match the number of text prompts. Note" " that this behavior is deprecated and will be removed in a version 1.0.0. Please make sure to update" " your script to pass as many initial images as text prompts to suppress this warning." ) deprecate("len(prompt) != len(image)", "1.0.0", deprecation_message, standard_warn=False) additional_image_per_prompt = batch_size // image_latents.shape[0] image_latents = torch.cat([image_latents] * additional_image_per_prompt, dim=0) elif batch_size > image_latents.shape[0] and batch_size % image_latents.shape[0] != 0: raise ValueError( f"Cannot duplicate `image` of batch size {image_latents.shape[0]} to {batch_size} text prompts." ) else: image_latents = torch.cat([image_latents], dim=0) if do_classifier_free_guidance: uncond_image_latents = torch.zeros_like(image_latents) image_latents = torch.cat([image_latents, image_latents, uncond_image_latents], dim=0) return image_latents @property def guidance_scale(self): return self._guidance_scale @property def image_guidance_scale(self): return self._image_guidance_scale @property def clip_skip(self): return self._clip_skip # here `guidance_scale` is defined analog to the guidance weight `w` of equation (2) # of the Imagen paper: https://arxiv.org/pdf/2205.11487.pdf . `guidance_scale = 1` # corresponds to doing no classifier free guidance. @property def do_classifier_free_guidance(self): return self.guidance_scale > 1.0 and self.image_guidance_scale >= 1.0 @property def joint_attention_kwargs(self): return self._joint_attention_kwargs @property def num_timesteps(self): return self._num_timesteps @property def interrupt(self): return self._interrupt @torch.no_grad() @replace_example_docstring(EXAMPLE_DOC_STRING) def __call__( self, prompt: Union[str, List[str]] = None, prompt_2: Optional[Union[str, List[str]]] = None, prompt_3: Optional[Union[str, List[str]]] = None, strength: float = 1.0, image: PipelineImageInput = None, height: Optional[int] = None, width: Optional[int] = None, num_inference_steps: int = 28, timesteps: List[int] = None, guidance_scale: float = 7.0, image_guidance_scale: float = 1.5, negative_prompt: Optional[Union[str, List[str]]] = None, negative_prompt_2: Optional[Union[str, List[str]]] = None, negative_prompt_3: Optional[Union[str, List[str]]] = None, num_images_per_prompt: Optional[int] = 1, 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, output_type: Optional[str] = "pil", return_dict: bool = True, joint_attention_kwargs: Optional[Dict[str, Any]] = None, clip_skip: Optional[int] = None, callback_on_step_end: Optional[Callable[[int, int, Dict], None]] = None, callback_on_step_end_tensor_inputs: List[str] = ["latents"], mask_img: Optional[PipelineImageInput] = None, **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 `tokenizer_2` and `text_encoder_2`. If not defined, `prompt` is will be used instead prompt_3 (`str` or `List[str]`, *optional*): The prompt or prompts to be sent to `tokenizer_3` and `text_encoder_3`. If not defined, `prompt` is will be used instead 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. 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. 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. 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 instead negative_prompt_3 (`str` or `List[str]`, *optional*): The prompt or prompts not to guide the image generation to be sent to `tokenizer_3` and `text_encoder_3`. If not defined, `negative_prompt` is used instead num_images_per_prompt (`int`, *optional*, defaults to 1): The number of images to generate per prompt. 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. 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. joint_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). 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. Examples: Returns: [`~pipelines.stable_diffusion_xl.StableDiffusionXLPipelineOutput`] or `tuple`: [`~pipelines.stable_diffusion_xl.StableDiffusionXLPipelineOutput`] if `return_dict` is True, otherwise a `tuple`. When returning a tuple, the first element is a list with the generated images. """ # height = height or self.default_sample_size * self.vae_scale_factor # width = width or self.default_sample_size * self.vae_scale_factor # 1. Check inputs. Raise error if not correct self.check_inputs( prompt, prompt_2, prompt_3, negative_prompt=negative_prompt, negative_prompt_2=negative_prompt_2, negative_prompt_3=negative_prompt_3, prompt_embeds=prompt_embeds, negative_prompt_embeds=negative_prompt_embeds, pooled_prompt_embeds=pooled_prompt_embeds, negative_pooled_prompt_embeds=negative_pooled_prompt_embeds, callback_on_step_end_tensor_inputs=callback_on_step_end_tensor_inputs, ) self._guidance_scale = guidance_scale self._image_guidance_scale = image_guidance_scale self._clip_skip = clip_skip self._joint_attention_kwargs = joint_attention_kwargs self._interrupt = False # 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 ( prompt_embeds, negative_prompt_embeds, pooled_prompt_embeds, negative_pooled_prompt_embeds, ) = self.encode_prompt( prompt=prompt, prompt_2=prompt_2, prompt_3=prompt_3, negative_prompt=negative_prompt, negative_prompt_2=negative_prompt_2, negative_prompt_3=negative_prompt_3, do_classifier_free_guidance=self.do_classifier_free_guidance, prompt_embeds=prompt_embeds, negative_prompt_embeds=negative_prompt_embeds, pooled_prompt_embeds=pooled_prompt_embeds, negative_pooled_prompt_embeds=negative_pooled_prompt_embeds, device=device, clip_skip=self.clip_skip, num_images_per_prompt=num_images_per_prompt, ) if self.do_classifier_free_guidance: # duplicate unconditional embeddings for each generation per prompt, using mps friendly method prompt_embeds = torch.cat([prompt_embeds, negative_prompt_embeds, negative_prompt_embeds], dim=0) # Similiarly pooled_prompt_embeds = torch.cat([pooled_prompt_embeds, negative_pooled_prompt_embeds, negative_pooled_prompt_embeds], dim=0) # if self.do_classifier_free_guidance: # prompt_embeds = torch.cat([negative_prompt_embeds, prompt_embeds], dim=0) # pooled_prompt_embeds = torch.cat([negative_pooled_prompt_embeds, pooled_prompt_embeds], dim=0) # 3. Preprocess image image = self.image_processor.preprocess(image) # 4. Prepare timesteps timesteps, num_inference_steps = retrieve_timesteps(self.scheduler, num_inference_steps, device, timesteps) timesteps, num_inference_steps = self.get_timesteps(num_inference_steps, strength, device) latent_timestep = timesteps[:1].repeat(batch_size * num_inference_steps) # timesteps, num_inference_steps = retrieve_timesteps(self.scheduler, num_inference_steps, device, timesteps) num_warmup_steps = max(len(timesteps) - num_inference_steps * self.scheduler.order, 0) self._num_timesteps = len(timesteps) # 5. Prepare Image latent image_latents = self.prepare_image_latents( image, batch_size, num_images_per_prompt, prompt_embeds.dtype, device, self.do_classifier_free_guidance, ) height, width = image_latents.shape[-2:] height = height * self.vae_scale_factor width = width * self.vae_scale_factor # 6. Prepare latent variables num_channels_latents = self.vae.config.latent_channels if latents is None: latents = self.prepare_latents( image, latent_timestep, batch_size, num_images_per_prompt, prompt_embeds.dtype, device, generator, ) else: return latents.to(device=device, dtype=prompt_embeds.dtype) # 7. Check that shapes of latents and image match the DIT in_channels num_channels_image = image_latents.shape[1] if mask_img is not None: mask_img = self.image_processor.preprocess(mask_img) mask_image_latents = self.prepare_image_latents( mask_img, batch_size, num_images_per_prompt, prompt_embeds.dtype, device, self.do_classifier_free_guidance, ) num_channels_image += mask_image_latents.shape[1] if num_channels_latents + num_channels_image != self.transformer.config.in_channels: raise ValueError( f"Incorrect configuration settings! The config of `pipeline.transformer`: {self.transformer.config} expects" f" {self.transformer.config.in_channels} but received `num_channels_latents`: {num_channels_latents} +" f" `num_channels_image`: {num_channels_image} " f" = {num_channels_latents+num_channels_image}. Please verify the config of" " `pipeline.transformer` or your `image` input." ) # 8. Denoising loop with self.progress_bar(total=num_inference_steps) as progress_bar: for i, t in enumerate(timesteps): if self.interrupt: continue # expand the latents if we are doing classifier free guidance latent_model_input = torch.cat([latents] * 3) if self.do_classifier_free_guidance else latents # broadcast to batch dimension in a way that's compatible with ONNX/Core ML timestep = t.expand(latent_model_input.shape[0]) scaled_latent_model_input = torch.cat([latent_model_input, image_latents], dim=1) if mask_img is not None: scaled_latent_model_input = torch.cat([scaled_latent_model_input, mask_image_latents], dim=1) # if "mask_index" in kwargs and kwargs['mask_index'] is not None: # mask_index = kwargs['mask_index'] # else: # mask_index = None noise_pred = self.transformer( hidden_states=scaled_latent_model_input, timestep=timestep, encoder_hidden_states=prompt_embeds, pooled_projections=pooled_prompt_embeds, joint_attention_kwargs=self.joint_attention_kwargs, return_dict=False, # mask_index= mask_index, )[0] # perform guidance if self.do_classifier_free_guidance: noise_pred_text, noise_pred_image, noise_pred_uncond = noise_pred.chunk(3) noise_pred = ( noise_pred_uncond + self.guidance_scale * (noise_pred_text - noise_pred_image) + self.image_guidance_scale * (noise_pred_image - noise_pred_uncond) ) # noise_pred_uncond, noise_pred_text = noise_pred.chunk(2) # neg, prompt # noise_pred = noise_pred_uncond + self.guidance_scale * (noise_pred_text - noise_pred_uncond) # compute the previous noisy sample x_t -> x_t-1 latents_dtype = latents.dtype latents = self.scheduler.step(noise_pred, t, latents, return_dict=False)[0] if latents.dtype != latents_dtype: if torch.backends.mps.is_available(): # some platforms (eg. apple mps) misbehave due to a pytorch bug: https://github.com/pytorch/pytorch/pull/99272 latents = latents.to(latents_dtype) 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) negative_pooled_prompt_embeds = callback_outputs.pop( "negative_pooled_prompt_embeds", negative_pooled_prompt_embeds ) image_latents = callback_outputs.pop("image_latents", image_latents) if mask_img is not None: mask_image_latents = callback_outputs.pop("mask_image_latents", mask_image_latents) # 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 XLA_AVAILABLE: xm.mark_step() if output_type == "latent": image = latents else: # latents = (latents / self.vae.config.scaling_factor) + self.vae.config.shift_factor latents = latents / self.vae.config.scaling_factor image = self.vae.decode(latents, return_dict=False)[0] image = self.image_processor.postprocess(image, output_type=output_type) # Offload all models self.maybe_free_model_hooks() if not return_dict: return (image,) return StableDiffusion3PipelineOutput(images=image)