diff --git a/examples/community/README.md b/examples/community/README.md index 0c4fd9aa82a3..9d2452e9177a 100644 --- a/examples/community/README.md +++ b/examples/community/README.md @@ -85,7 +85,7 @@ PIXART-α Controlnet pipeline | Implementation of the controlnet model for pixar | Stable Diffusion XL Attentive Eraser Pipeline |[[AAAI2025 Oral] Attentive Eraser](https://github.com/Anonym0u3/AttentiveEraser) is a novel tuning-free method that enhances object removal capabilities in pre-trained diffusion models.|[Stable Diffusion XL Attentive Eraser Pipeline](#stable-diffusion-xl-attentive-eraser-pipeline)|-|[Wenhao Sun](https://github.com/Anonym0u3) and [Benlei Cui](https://github.com/Benny079)| | Perturbed-Attention Guidance |StableDiffusionPAGPipeline is a modification of StableDiffusionPipeline to support Perturbed-Attention Guidance (PAG).|[Perturbed-Attention Guidance](#perturbed-attention-guidance)|[Notebook](https://github.com/huggingface/notebooks/blob/main/diffusers/perturbed_attention_guidance.ipynb)|[Hyoungwon Cho](https://github.com/HyoungwonCho)| | CogVideoX DDIM Inversion Pipeline | Implementation of DDIM inversion and guided attention-based editing denoising process on CogVideoX. | [CogVideoX DDIM Inversion Pipeline](#cogvideox-ddim-inversion-pipeline) | - | [LittleNyima](https://github.com/LittleNyima) | - +| FaithDiff Stable Diffusion XL Pipeline | Implementation of [(CVPR 2025) FaithDiff: Unleashing Diffusion Priors for Faithful Image Super-resolutionUnleashing Diffusion Priors for Faithful Image Super-resolution](https://arxiv.org/abs/2411.18824) - FaithDiff is a faithful image super-resolution method that leverages latent diffusion models by actively adapting the diffusion prior and jointly fine-tuning its components (encoder and diffusion model) with an alignment module to ensure high fidelity and structural consistency. | [FaithDiff Stable Diffusion XL Pipeline](#faithdiff-stable-diffusion-xl-pipeline) | [![Hugging Face Models](https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Models-blue)](https://huggingface.co/jychen9811/FaithDiff) | [Junyang Chen, Jinshan Pan, Jiangxin Dong, IMAG Lab, (Adapted by Eliseu Silva)](https://github.com/JyChen9811/FaithDiff) | To load a custom pipeline you just need to pass the `custom_pipeline` argument to `DiffusionPipeline`, as one of the files in `diffusers/examples/community`. Feel free to send a PR with your own pipelines, we will merge them quickly. ```py @@ -5334,3 +5334,103 @@ output = pipeline_for_inversion( pipeline.export_latents_to_video(output.inverse_latents[-1], "path/to/inverse_video.mp4", fps=8) pipeline.export_latents_to_video(output.recon_latents[-1], "path/to/recon_video.mp4", fps=8) ``` +# FaithDiff Stable Diffusion XL Pipeline + +[Project](https://jychen9811.github.io/FaithDiff_page/) / [GitHub](https://github.com/JyChen9811/FaithDiff/) + +This the implementation of the FaithDiff pipeline for SDXL, adapted to use the HuggingFace Diffusers. + +For more details see the project links above. + +## Example Usage + +This example upscale and restores a low-quality image. The input image has a resolution of 512x512 and will be upscaled at a scale of 2x, to a final resolution of 1024x1024. It is possible to upscale to a larger scale, but it is recommended that the input image be at least 1024x1024 in these cases. To upscale this image by 4x, for example, it would be recommended to re-input the result into a new 2x processing, thus performing progressive scaling. + +````py +import random +import numpy as np +import torch +from diffusers import DiffusionPipeline, AutoencoderKL, UniPCMultistepScheduler +from huggingface_hub import hf_hub_download +from diffusers.utils import load_image +from PIL import Image + +device = "cuda" +dtype = torch.float16 +MAX_SEED = np.iinfo(np.int32).max + +# Download weights for additional unet layers +model_file = hf_hub_download( + "jychen9811/FaithDiff", + filename="FaithDiff.bin", local_dir="./proc_data/faithdiff", local_dir_use_symlinks=False +) + +# Initialize the models and pipeline +vae = AutoencoderKL.from_pretrained("madebyollin/sdxl-vae-fp16-fix", torch_dtype=dtype) + +model_id = "SG161222/RealVisXL_V4.0" +pipe = DiffusionPipeline.from_pretrained( + model_id, + torch_dtype=dtype, + vae=vae, + unet=None, #<- Do not load with original model. + custom_pipeline="pipeline_faithdiff_stable_diffusion_xl", + use_safetensors=True, + variant="fp16", +).to(device) + +# Here we need use pipeline internal unet model +pipe.unet = pipe.unet_model.from_pretrained(model_id, subfolder="unet", variant="fp16", use_safetensors=True) + +# Load aditional layers to the model +pipe.unet.load_additional_layers(weight_path="proc_data/faithdiff/FaithDiff.bin", dtype=dtype) + +# Enable vae tiling +pipe.set_encoder_tile_settings() +pipe.enable_vae_tiling() + +# Optimization +pipe.enable_model_cpu_offload() + +# Set selected scheduler +pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config) + +#input params +prompt = "The image features a woman in her 55s with blonde hair and a white shirt, smiling at the camera. She appears to be in a good mood and is wearing a white scarf around her neck. " +upscale = 2 # scale here +start_point = "lr" # or "noise" +latent_tiled_overlap = 0.5 +latent_tiled_size = 1024 + +# Load image +lq_image = load_image("https://huggingface.co/datasets/DEVAIEXP/assets/resolve/main/woman.png") +original_height = lq_image.height +original_width = lq_image.width +print(f"Current resolution: H:{original_height} x W:{original_width}") + +width = original_width * int(upscale) +height = original_height * int(upscale) +print(f"Final resolution: H:{height} x W:{width}") + +# Restoration +image = lq_image.resize((width, height), Image.LANCZOS) +input_image, width_init, height_init, width_now, height_now = pipe.check_image_size(image) + +generator = torch.Generator(device=device).manual_seed(random.randint(0, MAX_SEED)) +gen_image = pipe(lr_img=input_image, + prompt = prompt, + num_inference_steps=20, + guidance_scale=5, + generator=generator, + start_point=start_point, + height = height_now, + width=width_now, + overlap=latent_tiled_overlap, + target_size=(latent_tiled_size, latent_tiled_size) + ).images[0] + +cropped_image = gen_image.crop((0, 0, width_init, height_init)) +cropped_image.save("data/result.png") +```` +### Result +[](https://imgsli.com/MzY1NzE2) \ No newline at end of file diff --git a/examples/community/pipeline_faithdiff_stable_diffusion_xl.py b/examples/community/pipeline_faithdiff_stable_diffusion_xl.py new file mode 100644 index 000000000000..d1d3d80b4a60 --- /dev/null +++ b/examples/community/pipeline_faithdiff_stable_diffusion_xl.py @@ -0,0 +1,2269 @@ +# Copyright 2025 Junyang Chen, Jinshan Pan, Jiangxin Dong, IMAG Lab Team +# 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. + +import copy +import inspect +from collections import OrderedDict +from dataclasses import dataclass +from typing import Any, Callable, Dict, List, Optional, Tuple, Union + +import cv2 +import numpy as np +import PIL.Image +import torch +import torch.nn as nn +import torch.nn.functional as F +import torch.utils.checkpoint +from transformers import CLIPTextModel, CLIPTextModelWithProjection, CLIPTokenizer + +from diffusers import UNet2DConditionModel as OriginalUNet2DConditionModel +from diffusers.configuration_utils import ConfigMixin, register_to_config +from diffusers.image_processor import PipelineImageInput, VaeImageProcessor +from diffusers.loaders import ( + FromSingleFileMixin, + IPAdapterMixin, + PeftAdapterMixin, + StableDiffusionXLLoraLoaderMixin, + TextualInversionLoaderMixin, + UNet2DConditionLoadersMixin, +) +from diffusers.models import AutoencoderKL +from diffusers.models.attention_processor import ( + AttnProcessor2_0, + FusedAttnProcessor2_0, + LoRAAttnProcessor2_0, + LoRAXFormersAttnProcessor, + XFormersAttnProcessor, +) +from diffusers.models.lora import adjust_lora_scale_text_encoder +from diffusers.models.unets.unet_2d_blocks import UNetMidBlock2D, get_down_block +from diffusers.pipelines.pipeline_utils import DiffusionPipeline, StableDiffusionMixin +from diffusers.pipelines.stable_diffusion_xl.pipeline_output import StableDiffusionXLPipelineOutput +from diffusers.schedulers import DDPMScheduler, KarrasDiffusionSchedulers +from diffusers.utils import ( + USE_PEFT_BACKEND, + deprecate, + is_invisible_watermark_available, + is_torch_version, + is_torch_xla_available, + logging, + replace_example_docstring, + scale_lora_layers, + unscale_lora_layers, +) +from diffusers.utils.outputs import BaseOutput +from diffusers.utils.torch_utils import randn_tensor + + +if is_invisible_watermark_available(): + from diffusers.pipelines.stable_diffusion_xl.watermark import StableDiffusionXLWatermarker + +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 random + >>> import numpy as np + >>> import torch + >>> from diffusers import DiffusionPipeline, AutoencoderKL, UniPCMultistepScheduler + >>> from huggingface_hub import hf_hub_download + >>> from diffusers.utils import load_image + >>> from PIL import Image + >>> + >>> device = "cuda" + >>> dtype = torch.float16 + >>> MAX_SEED = np.iinfo(np.int32).max + >>> + >>> # Download weights for additional unet layers + >>> model_file = hf_hub_download( + ... "jychen9811/FaithDiff", + ... filename="FaithDiff.bin", local_dir="./proc_data/faithdiff", local_dir_use_symlinks=False + ... ) + >>> + >>> # Initialize the models and pipeline + >>> vae = AutoencoderKL.from_pretrained("madebyollin/sdxl-vae-fp16-fix", torch_dtype=dtype) + >>> + >>> model_id = "SG161222/RealVisXL_V4.0" + >>> pipe = DiffusionPipeline.from_pretrained( + ... model_id, + ... torch_dtype=dtype, + ... vae=vae, + ... unet=None, #<- Do not load with original model. + ... custom_pipeline="mixture_tiling_sdxl", + ... use_safetensors=True, + ... variant="fp16", + ... ).to(device) + >>> + >>> # Here we need use pipeline internal unet model + >>> pipe.unet = pipe.unet_model.from_pretrained(model_id, subfolder="unet", variant="fp16", use_safetensors=True) + >>> + >>> # Load aditional layers to the model + >>> pipe.unet.load_additional_layers(weight_path="proc_data/faithdiff/FaithDiff.bin", dtype=dtype) + >>> + >>> # Enable vae tiling + >>> pipe.set_encoder_tile_settings() + >>> pipe.enable_vae_tiling() + >>> + >>> # Optimization + >>> pipe.enable_model_cpu_offload() + >>> + >>> # Set selected scheduler + >>> pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config) + >>> + >>> #input params + >>> prompt = "The image features a woman in her 55s with blonde hair and a white shirt, smiling at the camera. She appears to be in a good mood and is wearing a white scarf around her neck. " + >>> upscale = 2 # scale here + >>> start_point = "lr" # or "noise" + >>> latent_tiled_overlap = 0.5 + >>> latent_tiled_size = 1024 + >>> + >>> # Load image + >>> lq_image = load_image("https://huggingface.co/datasets/DEVAIEXP/assets/resolve/main/woman.png") + >>> original_height = lq_image.height + >>> original_width = lq_image.width + >>> print(f"Current resolution: H:{original_height} x W:{original_width}") + >>> + >>> width = original_width * int(upscale) + >>> height = original_height * int(upscale) + >>> print(f"Final resolution: H:{height} x W:{width}") + >>> + >>> # Restoration + >>> image = lq_image.resize((width, height), Image.LANCZOS) + >>> input_image, width_init, height_init, width_now, height_now = pipe.check_image_size(image) + >>> + >>> generator = torch.Generator(device=device).manual_seed(random.randint(0, MAX_SEED)) + >>> gen_image = pipe(lr_img=input_image, + ... prompt = prompt, + ... num_inference_steps=20, + ... guidance_scale=5, + ... generator=generator, + ... start_point=start_point, + ... height = height_now, + ... width=width_now, + ... overlap=latent_tiled_overlap, + ... target_size=(latent_tiled_size, latent_tiled_size) + ... ).images[0] + >>> + >>> cropped_image = gen_image.crop((0, 0, width_init, height_init)) + >>> cropped_image.save("data/result.png") + ``` +""" + + +def zero_module(module): + """Zero out the parameters of a module and return it.""" + for p in module.parameters(): + nn.init.zeros_(p) + return module + + +class Encoder(nn.Module): + """Encoder layer of a variational autoencoder that encodes input into a latent representation.""" + + def __init__( + self, + in_channels: int = 3, + out_channels: int = 4, + down_block_types: Tuple[str, ...] = ( + "DownEncoderBlock2D", + "DownEncoderBlock2D", + "DownEncoderBlock2D", + "DownEncoderBlock2D", + ), + block_out_channels: Tuple[int, ...] = (128, 256, 512, 512), + layers_per_block: int = 2, + norm_num_groups: int = 32, + act_fn: str = "silu", + double_z: bool = True, + mid_block_add_attention: bool = True, + ): + super().__init__() + self.layers_per_block = layers_per_block + + self.conv_in = nn.Conv2d( + in_channels, + block_out_channels[0], + kernel_size=3, + stride=1, + padding=1, + ) + + self.mid_block = None + self.down_blocks = nn.ModuleList([]) + self.use_rgb = False + self.down_block_type = down_block_types + self.block_out_channels = block_out_channels + + self.tile_sample_min_size = 1024 + self.tile_latent_min_size = int(self.tile_sample_min_size / 8) + self.tile_overlap_factor = 0.25 + self.use_tiling = False + + output_channel = block_out_channels[0] + for i, down_block_type in enumerate(down_block_types): + input_channel = output_channel + output_channel = block_out_channels[i] + is_final_block = i == len(block_out_channels) - 1 + + down_block = get_down_block( + down_block_type, + num_layers=self.layers_per_block, + in_channels=input_channel, + out_channels=output_channel, + add_downsample=not is_final_block, + resnet_eps=1e-6, + downsample_padding=0, + resnet_act_fn=act_fn, + resnet_groups=norm_num_groups, + attention_head_dim=output_channel, + temb_channels=None, + ) + self.down_blocks.append(down_block) + + self.mid_block = UNetMidBlock2D( + in_channels=block_out_channels[-1], + resnet_eps=1e-6, + resnet_act_fn=act_fn, + output_scale_factor=1, + resnet_time_scale_shift="default", + attention_head_dim=block_out_channels[-1], + resnet_groups=norm_num_groups, + temb_channels=None, + add_attention=mid_block_add_attention, + ) + + self.gradient_checkpointing = False + + def to_rgb_init(self): + """Initialize layers to convert features to RGB.""" + self.to_rgbs = nn.ModuleList([]) + self.use_rgb = True + for i, down_block_type in enumerate(self.down_block_type): + output_channel = self.block_out_channels[i] + self.to_rgbs.append(nn.Conv2d(output_channel, 3, kernel_size=3, padding=1)) + + def enable_tiling(self): + """Enable tiling for large inputs.""" + self.use_tiling = True + + def encode(self, sample: torch.FloatTensor) -> torch.FloatTensor: + """Encode the input tensor into a latent representation.""" + sample = self.conv_in(sample) + if self.training and self.gradient_checkpointing: + + def create_custom_forward(module): + def custom_forward(*inputs): + return module(*inputs) + + return custom_forward + + if is_torch_version(">=", "1.11.0"): + for down_block in self.down_blocks: + sample = torch.utils.checkpoint.checkpoint( + create_custom_forward(down_block), sample, use_reentrant=False + ) + sample = torch.utils.checkpoint.checkpoint( + create_custom_forward(self.mid_block), sample, use_reentrant=False + ) + else: + for down_block in self.down_blocks: + sample = torch.utils.checkpoint.checkpoint(create_custom_forward(down_block), sample) + sample = torch.utils.checkpoint.checkpoint(create_custom_forward(self.mid_block), sample) + return sample + else: + for down_block in self.down_blocks: + sample = down_block(sample) + sample = self.mid_block(sample) + return sample + + def blend_v(self, a: torch.Tensor, b: torch.Tensor, blend_extent: int) -> torch.Tensor: + """Blend two tensors vertically with a smooth transition.""" + blend_extent = min(a.shape[2], b.shape[2], blend_extent) + for y in range(blend_extent): + b[:, :, y, :] = a[:, :, -blend_extent + y, :] * (1 - y / blend_extent) + b[:, :, y, :] * (y / blend_extent) + return b + + def blend_h(self, a: torch.Tensor, b: torch.Tensor, blend_extent: int) -> torch.Tensor: + """Blend two tensors horizontally with a smooth transition.""" + blend_extent = min(a.shape[3], b.shape[3], blend_extent) + for x in range(blend_extent): + b[:, :, :, x] = a[:, :, :, -blend_extent + x] * (1 - x / blend_extent) + b[:, :, :, x] * (x / blend_extent) + return b + + def tiled_encode(self, x: torch.FloatTensor) -> torch.FloatTensor: + """Encode the input tensor using tiling for large inputs.""" + overlap_size = int(self.tile_sample_min_size * (1 - self.tile_overlap_factor)) + blend_extent = int(self.tile_latent_min_size * self.tile_overlap_factor) + row_limit = self.tile_latent_min_size - blend_extent + + rows = [] + for i in range(0, x.shape[2], overlap_size): + row = [] + for j in range(0, x.shape[3], overlap_size): + tile = x[:, :, i : i + self.tile_sample_min_size, j : j + self.tile_sample_min_size] + tile = self.encode(tile) + row.append(tile) + rows.append(row) + result_rows = [] + for i, row in enumerate(rows): + result_row = [] + for j, tile in enumerate(row): + if i > 0: + tile = self.blend_v(rows[i - 1][j], tile, blend_extent) + if j > 0: + tile = self.blend_h(row[j - 1], tile, blend_extent) + result_row.append(tile[:, :, :row_limit, :row_limit]) + result_rows.append(torch.cat(result_row, dim=3)) + + moments = torch.cat(result_rows, dim=2) + return moments + + def forward(self, sample: torch.FloatTensor) -> torch.FloatTensor: + """Forward pass of the encoder, using tiling if enabled for large inputs.""" + if self.use_tiling and ( + sample.shape[-1] > self.tile_latent_min_size or sample.shape[-2] > self.tile_latent_min_size + ): + return self.tiled_encode(sample) + return self.encode(sample) + + +class ControlNetConditioningEmbedding(nn.Module): + """A small network to preprocess conditioning inputs, inspired by ControlNet.""" + + def __init__(self, conditioning_embedding_channels: int, conditioning_channels: int = 4): + super().__init__() + self.conv_in = nn.Conv2d(conditioning_channels, conditioning_channels, kernel_size=3, padding=1) + self.norm_in = nn.GroupNorm(num_channels=conditioning_channels, num_groups=32, eps=1e-6) + self.conv_out = zero_module( + nn.Conv2d(conditioning_channels, conditioning_embedding_channels, kernel_size=3, padding=1) + ) + + def forward(self, conditioning): + """Process the conditioning input through the network.""" + conditioning = self.norm_in(conditioning) + embedding = self.conv_in(conditioning) + embedding = F.silu(embedding) + embedding = self.conv_out(embedding) + return embedding + + +class QuickGELU(nn.Module): + """A fast approximation of the GELU activation function.""" + + def forward(self, x: torch.Tensor): + """Apply the QuickGELU activation to the input tensor.""" + return x * torch.sigmoid(1.702 * x) + + +class LayerNorm(nn.LayerNorm): + """Subclass torch's LayerNorm to handle fp16.""" + + def forward(self, x: torch.Tensor): + """Apply LayerNorm and preserve the input dtype.""" + orig_type = x.dtype + ret = super().forward(x) + return ret.type(orig_type) + + +class ResidualAttentionBlock(nn.Module): + """A transformer-style block with self-attention and an MLP.""" + + def __init__(self, d_model: int, n_head: int, attn_mask: torch.Tensor = None): + super().__init__() + self.attn = nn.MultiheadAttention(d_model, n_head) + self.ln_1 = LayerNorm(d_model) + self.mlp = nn.Sequential( + OrderedDict( + [ + ("c_fc", nn.Linear(d_model, d_model * 2)), + ("gelu", QuickGELU()), + ("c_proj", nn.Linear(d_model * 2, d_model)), + ] + ) + ) + self.ln_2 = LayerNorm(d_model) + self.attn_mask = attn_mask + + def attention(self, x: torch.Tensor): + """Apply self-attention to the input tensor.""" + self.attn_mask = self.attn_mask.to(dtype=x.dtype, device=x.device) if self.attn_mask is not None else None + return self.attn(x, x, x, need_weights=False, attn_mask=self.attn_mask)[0] + + def forward(self, x: torch.Tensor): + """Forward pass through the residual attention block.""" + x = x + self.attention(self.ln_1(x)) + x = x + self.mlp(self.ln_2(x)) + return x + + +@dataclass +class UNet2DConditionOutput(BaseOutput): + """The output of UnifiedUNet2DConditionModel.""" + + sample: torch.FloatTensor = None + + +class UNet2DConditionModel(OriginalUNet2DConditionModel, ConfigMixin, UNet2DConditionLoadersMixin, PeftAdapterMixin): + """A unified 2D UNet model extending OriginalUNet2DConditionModel with custom functionality.""" + + _supports_gradient_checkpointing = True + + @register_to_config + def __init__( + self, + sample_size: Optional[int] = None, + in_channels: int = 4, + out_channels: int = 4, + center_input_sample: bool = False, + flip_sin_to_cos: bool = True, + freq_shift: int = 0, + down_block_types: Tuple[str] = ( + "CrossAttnDownBlock2D", + "CrossAttnDownBlock2D", + "CrossAttnDownBlock2D", + "DownBlock2D", + ), + mid_block_type: Optional[str] = "UNetMidBlock2DCrossAttn", + up_block_types: Tuple[str] = ("UpBlock2D", "CrossAttnUpBlock2D", "CrossAttnUpBlock2D", "CrossAttnUpBlock2D"), + only_cross_attention: Union[bool, Tuple[bool]] = False, + block_out_channels: Tuple[int] = (320, 640, 1280, 1280), + layers_per_block: Union[int, Tuple[int]] = 2, + downsample_padding: int = 1, + mid_block_scale_factor: float = 1, + dropout: float = 0.0, + act_fn: str = "silu", + norm_num_groups: Optional[int] = 32, + norm_eps: float = 1e-5, + cross_attention_dim: Union[int, Tuple[int]] = 1280, + transformer_layers_per_block: Union[int, Tuple[int], Tuple[Tuple]] = 1, + reverse_transformer_layers_per_block: Optional[Tuple[Tuple[int]]] = None, + encoder_hid_dim: Optional[int] = None, + encoder_hid_dim_type: Optional[str] = None, + attention_head_dim: Union[int, Tuple[int]] = 8, + num_attention_heads: Optional[Union[int, Tuple[int]]] = None, + dual_cross_attention: bool = False, + use_linear_projection: bool = False, + class_embed_type: Optional[str] = None, + addition_embed_type: Optional[str] = None, + addition_time_embed_dim: Optional[int] = None, + num_class_embeds: Optional[int] = None, + upcast_attention: bool = False, + resnet_time_scale_shift: str = "default", + resnet_skip_time_act: bool = False, + resnet_out_scale_factor: float = 1.0, + time_embedding_type: str = "positional", + time_embedding_dim: Optional[int] = None, + time_embedding_act_fn: Optional[str] = None, + timestep_post_act: Optional[str] = None, + time_cond_proj_dim: Optional[int] = None, + conv_in_kernel: int = 3, + conv_out_kernel: int = 3, + projection_class_embeddings_input_dim: Optional[int] = None, + attention_type: str = "default", + class_embeddings_concat: bool = False, + mid_block_only_cross_attention: Optional[bool] = None, + cross_attention_norm: Optional[str] = None, + addition_embed_type_num_heads: int = 64, + ): + """Initialize the UnifiedUNet2DConditionModel.""" + super().__init__( + sample_size=sample_size, + in_channels=in_channels, + out_channels=out_channels, + center_input_sample=center_input_sample, + flip_sin_to_cos=flip_sin_to_cos, + freq_shift=freq_shift, + down_block_types=down_block_types, + mid_block_type=mid_block_type, + up_block_types=up_block_types, + only_cross_attention=only_cross_attention, + block_out_channels=block_out_channels, + layers_per_block=layers_per_block, + downsample_padding=downsample_padding, + mid_block_scale_factor=mid_block_scale_factor, + dropout=dropout, + act_fn=act_fn, + norm_num_groups=norm_num_groups, + norm_eps=norm_eps, + cross_attention_dim=cross_attention_dim, + transformer_layers_per_block=transformer_layers_per_block, + reverse_transformer_layers_per_block=reverse_transformer_layers_per_block, + encoder_hid_dim=encoder_hid_dim, + encoder_hid_dim_type=encoder_hid_dim_type, + attention_head_dim=attention_head_dim, + num_attention_heads=num_attention_heads, + dual_cross_attention=dual_cross_attention, + use_linear_projection=use_linear_projection, + class_embed_type=class_embed_type, + addition_embed_type=addition_embed_type, + addition_time_embed_dim=addition_time_embed_dim, + num_class_embeds=num_class_embeds, + upcast_attention=upcast_attention, + resnet_time_scale_shift=resnet_time_scale_shift, + resnet_skip_time_act=resnet_skip_time_act, + resnet_out_scale_factor=resnet_out_scale_factor, + time_embedding_type=time_embedding_type, + time_embedding_dim=time_embedding_dim, + time_embedding_act_fn=time_embedding_act_fn, + timestep_post_act=timestep_post_act, + time_cond_proj_dim=time_cond_proj_dim, + conv_in_kernel=conv_in_kernel, + conv_out_kernel=conv_out_kernel, + projection_class_embeddings_input_dim=projection_class_embeddings_input_dim, + attention_type=attention_type, + class_embeddings_concat=class_embeddings_concat, + mid_block_only_cross_attention=mid_block_only_cross_attention, + cross_attention_norm=cross_attention_norm, + addition_embed_type_num_heads=addition_embed_type_num_heads, + ) + + # Additional attributes + self.denoise_encoder = None + self.information_transformer_layes = None + self.condition_embedding = None + self.agg_net = None + self.spatial_ch_projs = None + + def init_vae_encoder(self, dtype): + self.denoise_encoder = Encoder() + if dtype is not None: + self.denoise_encoder.dtype = dtype + + def init_information_transformer_layes(self): + num_trans_channel = 640 + num_trans_head = 8 + num_trans_layer = 2 + num_proj_channel = 320 + self.information_transformer_layes = nn.Sequential( + *[ResidualAttentionBlock(num_trans_channel, num_trans_head) for _ in range(num_trans_layer)] + ) + self.spatial_ch_projs = zero_module(nn.Linear(num_trans_channel, num_proj_channel)) + + def init_ControlNetConditioningEmbedding(self, channel=512): + self.condition_embedding = ControlNetConditioningEmbedding(320, channel) + + def init_extra_weights(self): + self.agg_net = nn.ModuleList() + + def load_additional_layers( + self, dtype: Optional[torch.dtype] = torch.float16, channel: int = 512, weight_path: Optional[str] = None + ): + """Load additional layers and weights from a file. + + Args: + weight_path (str): Path to the weight file. + dtype (torch.dtype, optional): Data type for the loaded weights. Defaults to torch.float16. + channel (int): Conditioning embedding channel out size. Defaults 512. + """ + if self.denoise_encoder is None: + self.init_vae_encoder(dtype) + + if self.information_transformer_layes is None: + self.init_information_transformer_layes() + + if self.condition_embedding is None: + self.init_ControlNetConditioningEmbedding(channel) + + if self.agg_net is None: + self.init_extra_weights() + + # Load weights if provided + if weight_path is not None: + state_dict = torch.load(weight_path, weights_only=False) + self.load_state_dict(state_dict, strict=True) + + # Move all modules to the same device and dtype as the model + device = next(self.parameters()).device + if dtype is not None or device is not None: + self.to(device=device, dtype=dtype or next(self.parameters()).dtype) + + def to(self, *args, **kwargs): + """Override to() to move all additional modules to the same device and dtype.""" + super().to(*args, **kwargs) + for module in [ + self.denoise_encoder, + self.information_transformer_layes, + self.condition_embedding, + self.agg_net, + self.spatial_ch_projs, + ]: + if module is not None: + module.to(*args, **kwargs) + return self + + def load_state_dict(self, state_dict, strict=True): + """Load state dictionary into the model. + + Args: + state_dict (dict): State dictionary to load. + strict (bool, optional): Whether to strictly enforce that all keys match. Defaults to True. + """ + core_dict = {} + additional_dicts = { + "denoise_encoder": {}, + "information_transformer_layes": {}, + "condition_embedding": {}, + "agg_net": {}, + "spatial_ch_projs": {}, + } + + for key, value in state_dict.items(): + if key.startswith("denoise_encoder."): + additional_dicts["denoise_encoder"][key[len("denoise_encoder.") :]] = value + elif key.startswith("information_transformer_layes."): + additional_dicts["information_transformer_layes"][key[len("information_transformer_layes.") :]] = value + elif key.startswith("condition_embedding."): + additional_dicts["condition_embedding"][key[len("condition_embedding.") :]] = value + elif key.startswith("agg_net."): + additional_dicts["agg_net"][key[len("agg_net.") :]] = value + elif key.startswith("spatial_ch_projs."): + additional_dicts["spatial_ch_projs"][key[len("spatial_ch_projs.") :]] = value + else: + core_dict[key] = value + + super().load_state_dict(core_dict, strict=False) + for module_name, module_dict in additional_dicts.items(): + module = getattr(self, module_name, None) + if module is not None and module_dict: + module.load_state_dict(module_dict, strict=strict) + + def forward( + self, + sample: torch.FloatTensor, + timestep: Union[torch.Tensor, float, int], + encoder_hidden_states: torch.Tensor, + class_labels: Optional[torch.Tensor] = None, + timestep_cond: Optional[torch.Tensor] = None, + attention_mask: Optional[torch.Tensor] = None, + cross_attention_kwargs: Optional[Dict[str, Any]] = None, + added_cond_kwargs: Optional[Dict[str, torch.Tensor]] = None, + down_block_additional_residuals: Optional[Tuple[torch.Tensor]] = None, + mid_block_additional_residual: Optional[torch.Tensor] = None, + down_intrablock_additional_residuals: Optional[Tuple[torch.Tensor]] = None, + encoder_attention_mask: Optional[torch.Tensor] = None, + input_embedding: Optional[torch.Tensor] = None, + add_sample: bool = True, + return_dict: bool = True, + use_condition_embedding: bool = True, + ) -> Union[UNet2DConditionOutput, Tuple]: + """Forward pass prioritizing the original modified implementation. + + Args: + sample (torch.FloatTensor): The noisy input tensor with shape `(batch, channel, height, width)`. + timestep (Union[torch.Tensor, float, int]): The number of timesteps to denoise an input. + encoder_hidden_states (torch.Tensor): The encoder hidden states with shape `(batch, sequence_length, feature_dim)`. + class_labels (torch.Tensor, optional): Optional class labels for conditioning. + timestep_cond (torch.Tensor, optional): Conditional embeddings for timestep. + attention_mask (torch.Tensor, optional): An attention mask of shape `(batch, key_tokens)`. + cross_attention_kwargs (Dict[str, Any], optional): A kwargs dictionary for the AttentionProcessor. + added_cond_kwargs (Dict[str, torch.Tensor], optional): Additional embeddings to add to the UNet blocks. + down_block_additional_residuals (Tuple[torch.Tensor], optional): Residuals for down UNet blocks. + mid_block_additional_residual (torch.Tensor, optional): Residual for the middle UNet block. + down_intrablock_additional_residuals (Tuple[torch.Tensor], optional): Additional residuals within down blocks. + encoder_attention_mask (torch.Tensor, optional): A cross-attention mask of shape `(batch, sequence_length)`. + input_embedding (torch.Tensor, optional): Additional input embedding for preprocessing. + add_sample (bool): Whether to add the sample to the processed embedding. Defaults to True. + return_dict (bool): Whether to return a UNet2DConditionOutput. Defaults to True. + use_condition_embedding (bool): Whether to use the condition embedding. Defaults to True. + + Returns: + Union[UNet2DConditionOutput, Tuple]: The processed sample tensor, either as a UNet2DConditionOutput or tuple. + """ + default_overall_up_factor = 2**self.num_upsamplers + forward_upsample_size = False + upsample_size = None + + for dim in sample.shape[-2:]: + if dim % default_overall_up_factor != 0: + forward_upsample_size = True + break + + if attention_mask is not None: + attention_mask = (1 - attention_mask.to(sample.dtype)) * -10000.0 + attention_mask = attention_mask.unsqueeze(1) + + if encoder_attention_mask is not None: + encoder_attention_mask = (1 - encoder_attention_mask.to(sample.dtype)) * -10000.0 + encoder_attention_mask = encoder_attention_mask.unsqueeze(1) + + if self.config.center_input_sample: + sample = 2 * sample - 1.0 + + # 1. time + t_emb = self.get_time_embed(sample=sample, timestep=timestep) + emb = self.time_embedding(t_emb, timestep_cond) + aug_emb = None + + class_emb = self.get_class_embed(sample=sample, class_labels=class_labels) + if class_emb is not None: + if self.config.class_embeddings_concat: + emb = torch.cat([emb, class_emb], dim=-1) + else: + emb = emb + class_emb + + aug_emb = self.get_aug_embed( + emb=emb, encoder_hidden_states=encoder_hidden_states, added_cond_kwargs=added_cond_kwargs + ) + if self.config.addition_embed_type == "image_hint": + aug_emb, hint = aug_emb + sample = torch.cat([sample, hint], dim=1) + + emb = emb + aug_emb if aug_emb is not None else emb + + if self.time_embed_act is not None: + emb = self.time_embed_act(emb) + + encoder_hidden_states = self.process_encoder_hidden_states( + encoder_hidden_states=encoder_hidden_states, added_cond_kwargs=added_cond_kwargs + ) + + # 2. pre-process (following the original modified logic) + sample = self.conv_in(sample) # [B, 4, H, W] -> [B, 320, H, W] + if ( + input_embedding is not None + and self.condition_embedding is not None + and self.information_transformer_layes is not None + ): + if use_condition_embedding: + input_embedding = self.condition_embedding(input_embedding) # [B, 320, H, W] + batch_size, channel, height, width = input_embedding.shape + concat_feat = ( + torch.cat([sample, input_embedding], dim=1) + .view(batch_size, 2 * channel, height * width) + .transpose(1, 2) + ) + concat_feat = self.information_transformer_layes(concat_feat) + feat_alpha = self.spatial_ch_projs(concat_feat).transpose(1, 2).view(batch_size, channel, height, width) + sample = sample + feat_alpha if add_sample else feat_alpha # Update sample as in the original version + + # 2.5 GLIGEN position net (kept from the original version) + if cross_attention_kwargs is not None and cross_attention_kwargs.get("gligen", None) is not None: + cross_attention_kwargs = cross_attention_kwargs.copy() + gligen_args = cross_attention_kwargs.pop("gligen") + cross_attention_kwargs["gligen"] = {"objs": self.position_net(**gligen_args)} + + # 3. down (continues the standard flow) + if cross_attention_kwargs is not None: + cross_attention_kwargs = cross_attention_kwargs.copy() + lora_scale = cross_attention_kwargs.pop("scale", 1.0) + else: + lora_scale = 1.0 + + if USE_PEFT_BACKEND: + scale_lora_layers(self, lora_scale) + + is_controlnet = mid_block_additional_residual is not None and down_block_additional_residuals is not None + is_adapter = down_intrablock_additional_residuals is not None + if not is_adapter and mid_block_additional_residual is None and down_block_additional_residuals is not None: + deprecate( + "T2I should not use down_block_additional_residuals", + "1.3.0", + "Passing intrablock residual connections with `down_block_additional_residuals` is deprecated \ + and will be removed in diffusers 1.3.0. `down_block_additional_residuals` should only be used \ + for ControlNet. Please make sure use `down_intrablock_additional_residuals` instead. ", + standard_warn=False, + ) + down_intrablock_additional_residuals = down_block_additional_residuals + is_adapter = True + + down_block_res_samples = (sample,) + for downsample_block in self.down_blocks: + if hasattr(downsample_block, "has_cross_attention") and downsample_block.has_cross_attention: + additional_residuals = {} + if is_adapter and len(down_intrablock_additional_residuals) > 0: + additional_residuals["additional_residuals"] = down_intrablock_additional_residuals.pop(0) + sample, res_samples = downsample_block( + hidden_states=sample, + temb=emb, + encoder_hidden_states=encoder_hidden_states, + attention_mask=attention_mask, + cross_attention_kwargs=cross_attention_kwargs, + encoder_attention_mask=encoder_attention_mask, + **additional_residuals, + ) + else: + sample, res_samples = downsample_block(hidden_states=sample, temb=emb) + if is_adapter and len(down_intrablock_additional_residuals) > 0: + sample += down_intrablock_additional_residuals.pop(0) + down_block_res_samples += res_samples + + if is_controlnet: + new_down_block_res_samples = () + for down_block_res_sample, down_block_additional_residual in zip( + down_block_res_samples, down_block_additional_residuals + ): + down_block_res_sample = down_block_res_sample + down_block_additional_residual + new_down_block_res_samples = new_down_block_res_samples + (down_block_res_sample,) + down_block_res_samples = new_down_block_res_samples + + # 4. mid + if self.mid_block is not None: + if hasattr(self.mid_block, "has_cross_attention") and self.mid_block.has_cross_attention: + sample = self.mid_block( + sample, + emb, + encoder_hidden_states=encoder_hidden_states, + attention_mask=attention_mask, + cross_attention_kwargs=cross_attention_kwargs, + encoder_attention_mask=encoder_attention_mask, + ) + else: + sample = self.mid_block(sample, emb) + if ( + is_adapter + and len(down_intrablock_additional_residuals) > 0 + and sample.shape == down_intrablock_additional_residuals[0].shape + ): + sample += down_intrablock_additional_residuals.pop(0) + + if is_controlnet: + sample = sample + mid_block_additional_residual + + # 5. up + for i, upsample_block in enumerate(self.up_blocks): + is_final_block = i == len(self.up_blocks) - 1 + res_samples = down_block_res_samples[-len(upsample_block.resnets) :] + down_block_res_samples = down_block_res_samples[: -len(upsample_block.resnets)] + if not is_final_block and forward_upsample_size: + upsample_size = down_block_res_samples[-1].shape[2:] + if hasattr(upsample_block, "has_cross_attention") and upsample_block.has_cross_attention: + sample = upsample_block( + hidden_states=sample, + temb=emb, + res_hidden_states_tuple=res_samples, + encoder_hidden_states=encoder_hidden_states, + cross_attention_kwargs=cross_attention_kwargs, + upsample_size=upsample_size, + attention_mask=attention_mask, + encoder_attention_mask=encoder_attention_mask, + ) + else: + sample = upsample_block( + hidden_states=sample, + temb=emb, + res_hidden_states_tuple=res_samples, + upsample_size=upsample_size, + ) + + # 6. post-process + if self.conv_norm_out: + sample = self.conv_norm_out(sample) + sample = self.conv_act(sample) + sample = self.conv_out(sample) + + if USE_PEFT_BACKEND: + unscale_lora_layers(self, lora_scale) + + if not return_dict: + return (sample,) + return UNet2DConditionOutput(sample=sample) + + +class LocalAttention: + """A class to handle local attention by splitting tensors into overlapping grids for processing.""" + + def __init__(self, kernel_size=None, overlap=0.5): + """Initialize the LocalAttention module. + + Args: + kernel_size (tuple[int, int], optional): Size of the grid (height, width). Defaults to None. + overlap (float): Overlap factor between adjacent grids (0.0 to 1.0). Defaults to 0.5. + """ + super().__init__() + self.kernel_size = kernel_size + self.overlap = overlap + + def grids_list(self, x): + """Split the input tensor into a list of non-overlapping grid patches. + + Args: + x (torch.Tensor): Input tensor of shape (batch, channels, height, width). + + Returns: + list[torch.Tensor]: List of tensor patches. + """ + b, c, h, w = x.shape + self.original_size = (b, c, h, w) + assert b == 1 + k1, k2 = self.kernel_size + if h < k1: + k1 = h + if w < k2: + k2 = w + num_row = (h - 1) // k1 + 1 + num_col = (w - 1) // k2 + 1 + self.nr = num_row + self.nc = num_col + + import math + + step_j = k2 if num_col == 1 else math.ceil(k2 * self.overlap) + step_i = k1 if num_row == 1 else math.ceil(k1 * self.overlap) + parts = [] + idxes = [] + i = 0 + last_i = False + while i < h and not last_i: + j = 0 + if i + k1 >= h: + i = h - k1 + last_i = True + last_j = False + while j < w and not last_j: + if j + k2 >= w: + j = w - k2 + last_j = True + parts.append(x[:, :, i : i + k1, j : j + k2]) + idxes.append({"i": i, "j": j}) + j = j + step_j + i = i + step_i + return parts + + def grids(self, x): + """Split the input tensor into overlapping grid patches and concatenate them. + + Args: + x (torch.Tensor): Input tensor of shape (batch, channels, height, width). + + Returns: + torch.Tensor: Concatenated tensor of all grid patches. + """ + b, c, h, w = x.shape + self.original_size = (b, c, h, w) + assert b == 1 + k1, k2 = self.kernel_size + if h < k1: + k1 = h + if w < k2: + k2 = w + self.tile_weights = self._gaussian_weights(k2, k1) + num_row = (h - 1) // k1 + 1 + num_col = (w - 1) // k2 + 1 + self.nr = num_row + self.nc = num_col + + import math + + step_j = k2 if num_col == 1 else math.ceil(k2 * self.overlap) + step_i = k1 if num_row == 1 else math.ceil(k1 * self.overlap) + parts = [] + idxes = [] + i = 0 + last_i = False + while i < h and not last_i: + j = 0 + if i + k1 >= h: + i = h - k1 + last_i = True + last_j = False + while j < w and not last_j: + if j + k2 >= w: + j = w - k2 + last_j = True + parts.append(x[:, :, i : i + k1, j : j + k2]) + idxes.append({"i": i, "j": j}) + j = j + step_j + i = i + step_i + self.idxes = idxes + return torch.cat(parts, dim=0) + + def _gaussian_weights(self, tile_width, tile_height): + """Generate a Gaussian weight mask for tile contributions. + + Args: + tile_width (int): Width of the tile. + tile_height (int): Height of the tile. + + Returns: + torch.Tensor: Gaussian weight tensor of shape (channels, height, width). + """ + import numpy as np + from numpy import exp, pi, sqrt + + latent_width = tile_width + latent_height = tile_height + var = 0.01 + midpoint = (latent_width - 1) / 2 + x_probs = [ + exp(-(x - midpoint) * (x - midpoint) / (latent_width * latent_width) / (2 * var)) / sqrt(2 * pi * var) + for x in range(latent_width) + ] + midpoint = latent_height / 2 + y_probs = [ + exp(-(y - midpoint) * (y - midpoint) / (latent_height * latent_height) / (2 * var)) / sqrt(2 * pi * var) + for y in range(latent_height) + ] + weights = np.outer(y_probs, x_probs) + return torch.tile(torch.tensor(weights, device=torch.device("cuda")), (4, 1, 1)) + + def grids_inverse(self, outs): + """Reconstruct the original tensor from processed grid patches with overlap blending. + + Args: + outs (torch.Tensor): Processed grid patches. + + Returns: + torch.Tensor: Reconstructed tensor of original size. + """ + preds = torch.zeros(self.original_size).to(outs.device) + b, c, h, w = self.original_size + count_mt = torch.zeros((b, 4, h, w)).to(outs.device) + k1, k2 = self.kernel_size + + for cnt, each_idx in enumerate(self.idxes): + i = each_idx["i"] + j = each_idx["j"] + preds[0, :, i : i + k1, j : j + k2] += outs[cnt, :, :, :] * self.tile_weights + count_mt[0, :, i : i + k1, j : j + k2] += self.tile_weights + + del outs + torch.cuda.empty_cache() + return preds / count_mt + + def _pad(self, x): + """Pad the input tensor to align with kernel size. + + Args: + x (torch.Tensor): Input tensor of shape (batch, channels, height, width). + + Returns: + tuple: Padded tensor and padding values. + """ + b, c, h, w = x.shape + k1, k2 = self.kernel_size + mod_pad_h = (k1 - h % k1) % k1 + mod_pad_w = (k2 - w % k2) % k2 + pad = (mod_pad_w // 2, mod_pad_w - mod_pad_w // 2, mod_pad_h // 2, mod_pad_h - mod_pad_h // 2) + x = F.pad(x, pad, "reflect") + return x, pad + + def forward(self, x): + """Apply local attention by splitting into grids and reconstructing. + + Args: + x (torch.Tensor): Input tensor of shape (batch, channels, height, width). + + Returns: + torch.Tensor: Processed tensor of original size. + """ + b, c, h, w = x.shape + qkv = self.grids(x) + out = self.grids_inverse(qkv) + return out + + +# 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 + + Args: + noise_cfg (torch.Tensor): Noise configuration tensor. + noise_pred_text (torch.Tensor): Predicted noise from text-conditioned model. + guidance_rescale (float): Rescaling factor for guidance. Defaults to 0.0. + + Returns: + torch.Tensor: Rescaled noise configuration. + """ + 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_img2img.retrieve_latents +def retrieve_latents( + encoder_output: torch.Tensor, generator: Optional[torch.Generator] = None, sample_mode: str = "sample" +): + """Retrieve latents from an encoder output. + + Args: + encoder_output (torch.Tensor): Output from an encoder (e.g., VAE). + generator (torch.Generator, optional): Random generator for sampling. Defaults to None. + sample_mode (str): Sampling mode ("sample" or "argmax"). Defaults to "sample". + + Returns: + torch.Tensor: Retrieved latent tensor. + """ + if hasattr(encoder_output, "latent_dist") and sample_mode == "sample": + return encoder_output.latent_dist.sample(generator) + elif hasattr(encoder_output, "latent_dist") and sample_mode == "argmax": + return encoder_output.latent_dist.mode() + elif hasattr(encoder_output, "latents"): + return encoder_output.latents + else: + raise AttributeError("Could not access latents of provided encoder_output") + + +# 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 + + +class FaithDiffStableDiffusionXLPipeline( + DiffusionPipeline, + StableDiffusionMixin, + FromSingleFileMixin, + StableDiffusionXLLoraLoaderMixin, + TextualInversionLoaderMixin, + IPAdapterMixin, +): + r""" + Pipeline for text-to-image generation using Stable Diffusion XL. + + This model inherits from [`DiffusionPipeline`]. Check the superclass documentation for the generic methods the + library implements for all the pipelines (such as downloading or saving, running on a particular device, etc.) + + The pipeline also inherits the following loading methods: + - [`~loaders.TextualInversionLoaderMixin.load_textual_inversion`] for loading textual inversion embeddings + - [`~loaders.FromSingleFileMixin.from_single_file`] for loading `.ckpt` files + - [`~loaders.StableDiffusionXLLoraLoaderMixin.load_lora_weights`] for loading LoRA weights + - [`~loaders.StableDiffusionXLLoraLoaderMixin.save_lora_weights`] for saving LoRA weights + - [`~loaders.IPAdapterMixin.load_ip_adapter`] for loading IP Adapters + + Args: + vae ([`AutoencoderKL`]): + Variational Auto-Encoder (VAE) Model to encode and decode images to and from latent representations. + text_encoder ([`CLIPTextModel`]): + Frozen text-encoder. Stable Diffusion XL uses the text portion of + [CLIP](https://huggingface.co/docs/transformers/model_doc/clip#transformers.CLIPTextModel), specifically + the [clip-vit-large-patch14](https://huggingface.co/openai/clip-vit-large-patch14) variant. + text_encoder_2 ([` CLIPTextModelWithProjection`]): + Second frozen text-encoder. Stable Diffusion XL uses the text and pool portion of + [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. + 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). + unet ([`UNet2DConditionModel`]): Conditional U-Net architecture to denoise the encoded image latents. + scheduler ([`SchedulerMixin`]): + A scheduler to be used in combination with `unet` to denoise the encoded image latents. Can be one of + [`DDIMScheduler`], [`LMSDiscreteScheduler`], or [`PNDMScheduler`]. + force_zeros_for_empty_prompt (`bool`, *optional*, defaults to `"True"`): + Whether the negative prompt embeddings shall be forced to always be set to 0. Also see the config of + `stabilityai/stable-diffusion-xl-base-1-0`. + add_watermarker (`bool`, *optional*): + Whether to use the [invisible_watermark library](https://github.com/ShieldMnt/invisible-watermark/) to + watermark output images. If not defined, it will default to True if the package is installed, otherwise no + watermarker will be used. + """ + + unet_model = UNet2DConditionModel + model_cpu_offload_seq = "text_encoder->text_encoder_2->unet->vae" + _optional_components = ["tokenizer", "tokenizer_2", "text_encoder", "text_encoder_2", "feature_extractor", "unet"] + _callback_tensor_inputs = [ + "latents", + "prompt_embeds", + "negative_prompt_embeds", + "add_text_embeds", + "add_time_ids", + "negative_pooled_prompt_embeds", + "negative_add_time_ids", + ] + + def __init__( + self, + vae: AutoencoderKL, + text_encoder: CLIPTextModel, + text_encoder_2: CLIPTextModelWithProjection, + tokenizer: CLIPTokenizer, + tokenizer_2: CLIPTokenizer, + unet: OriginalUNet2DConditionModel, + scheduler: KarrasDiffusionSchedulers, + force_zeros_for_empty_prompt: bool = True, + add_watermarker: Optional[bool] = None, + ): + super().__init__() + + self.register_modules( + vae=vae, + text_encoder=text_encoder, + text_encoder_2=text_encoder_2, + tokenizer=tokenizer, + tokenizer_2=tokenizer_2, + unet=unet, + scheduler=scheduler, + ) + self.register_to_config(force_zeros_for_empty_prompt=force_zeros_for_empty_prompt) + self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1) + self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor) + self.DDPMScheduler = DDPMScheduler.from_config(self.scheduler.config, subfolder="scheduler") + self.default_sample_size = self.unet.config.sample_size if unet is not None else 128 + add_watermarker = add_watermarker if add_watermarker is not None else is_invisible_watermark_available() + + if add_watermarker: + self.watermark = StableDiffusionXLWatermarker() + else: + self.watermark = None + + def encode_prompt( + self, + prompt: str, + prompt_2: Optional[str] = None, + device: Optional[torch.device] = None, + num_images_per_prompt: int = 1, + do_classifier_free_guidance: bool = True, + negative_prompt: Optional[str] = None, + negative_prompt_2: Optional[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, + lora_scale: Optional[float] = None, + clip_skip: Optional[int] = None, + ): + r""" + Encodes the prompt into text encoder hidden states. + + 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 both 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 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. + lora_scale (`float`, *optional*): + A lora scale that will be applied to all LoRA layers of the text encoder if LoRA layers are loaded. + 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 = "cuda" # device or self._execution_device + + # set lora scale so that monkey patched LoRA + # function of text encoder can correctly access it + if lora_scale is not None and isinstance(self, StableDiffusionXLLoraLoaderMixin): + self._lora_scale = lora_scale + + # dynamically adjust the LoRA scale + if self.text_encoder is not None: + if not USE_PEFT_BACKEND: + adjust_lora_scale_text_encoder(self.text_encoder, lora_scale) + else: + scale_lora_layers(self.text_encoder, lora_scale) + + if self.text_encoder_2 is not None: + if not USE_PEFT_BACKEND: + adjust_lora_scale_text_encoder(self.text_encoder_2, lora_scale) + else: + scale_lora_layers(self.text_encoder_2, lora_scale) + + prompt = [prompt] if isinstance(prompt, str) else prompt + + if prompt is not None: + batch_size = len(prompt) + else: + batch_size = prompt_embeds.shape[0] + + # Define tokenizers and text encoders + tokenizers = [self.tokenizer, self.tokenizer_2] if self.tokenizer is not None else [self.tokenizer_2] + text_encoders = ( + [self.text_encoder, self.text_encoder_2] if self.text_encoder is not None else [self.text_encoder_2] + ) + dtype = text_encoders[0].dtype + if prompt_embeds is None: + prompt_2 = prompt_2 or prompt + prompt_2 = [prompt_2] if isinstance(prompt_2, str) else prompt_2 + + # textual inversion: process multi-vector tokens if necessary + prompt_embeds_list = [] + prompts = [prompt, prompt_2] + for prompt, tokenizer, text_encoder in zip(prompts, tokenizers, text_encoders): + if isinstance(self, TextualInversionLoaderMixin): + prompt = self.maybe_convert_prompt(prompt, tokenizer) + + text_inputs = tokenizer( + prompt, + padding="max_length", + max_length=tokenizer.model_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[:, tokenizer.model_max_length - 1 : -1]) + logger.warning( + "The following part of your input was truncated because CLIP can only handle sequences up to" + f" {tokenizer.model_max_length} tokens: {removed_text}" + ) + text_encoder = text_encoder.to(dtype) + prompt_embeds = text_encoder(text_input_ids.to(device), output_hidden_states=True) + + # We are only ALWAYS interested in the pooled output of the final text encoder + pooled_prompt_embeds = prompt_embeds[0] + if clip_skip is None: + prompt_embeds = prompt_embeds.hidden_states[-2] + else: + # "2" because SDXL always indexes from the penultimate layer. + prompt_embeds = prompt_embeds.hidden_states[-(clip_skip + 2)] + + prompt_embeds_list.append(prompt_embeds) + + prompt_embeds = torch.concat(prompt_embeds_list, dim=-1) + + # get unconditional embeddings for classifier free guidance + zero_out_negative_prompt = negative_prompt is None and self.config.force_zeros_for_empty_prompt + if do_classifier_free_guidance and negative_prompt_embeds is None and zero_out_negative_prompt: + negative_prompt_embeds = torch.zeros_like(prompt_embeds) + negative_pooled_prompt_embeds = torch.zeros_like(pooled_prompt_embeds) + elif do_classifier_free_guidance and negative_prompt_embeds is None: + negative_prompt = negative_prompt or "" + negative_prompt_2 = negative_prompt_2 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 + ) + + uncond_tokens: List[str] + 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`." + ) + else: + uncond_tokens = [negative_prompt, negative_prompt_2] + + negative_prompt_embeds_list = [] + for negative_prompt, tokenizer, text_encoder in zip(uncond_tokens, tokenizers, text_encoders): + if isinstance(self, TextualInversionLoaderMixin): + negative_prompt = self.maybe_convert_prompt(negative_prompt, tokenizer) + + max_length = prompt_embeds.shape[1] + uncond_input = tokenizer( + negative_prompt, + padding="max_length", + max_length=max_length, + truncation=True, + return_tensors="pt", + ) + + negative_prompt_embeds = text_encoder( + uncond_input.input_ids.to(device), + output_hidden_states=True, + ) + # We are only ALWAYS interested in the pooled output of the final text encoder + negative_pooled_prompt_embeds = negative_prompt_embeds[0] + negative_prompt_embeds = negative_prompt_embeds.hidden_states[-2] + + negative_prompt_embeds_list.append(negative_prompt_embeds) + + negative_prompt_embeds = torch.concat(negative_prompt_embeds_list, dim=-1) + + if self.text_encoder_2 is not None: + prompt_embeds = prompt_embeds.to(dtype=self.text_encoder_2.dtype, device=device) + else: + prompt_embeds = prompt_embeds.to(dtype=self.unet.dtype, device=device) + + bs_embed, 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(bs_embed * num_images_per_prompt, seq_len, -1) + + if do_classifier_free_guidance: + # duplicate unconditional embeddings for each generation per prompt, using mps friendly method + seq_len = negative_prompt_embeds.shape[1] + + if self.text_encoder_2 is not None: + negative_prompt_embeds = negative_prompt_embeds.to(dtype=self.text_encoder_2.dtype, device=device) + else: + negative_prompt_embeds = negative_prompt_embeds.to(dtype=self.unet.dtype, device=device) + + negative_prompt_embeds = negative_prompt_embeds.repeat(1, num_images_per_prompt, 1) + negative_prompt_embeds = negative_prompt_embeds.view(batch_size * num_images_per_prompt, seq_len, -1) + + pooled_prompt_embeds = pooled_prompt_embeds.repeat(1, num_images_per_prompt).view( + bs_embed * num_images_per_prompt, -1 + ) + if do_classifier_free_guidance: + negative_pooled_prompt_embeds = negative_pooled_prompt_embeds.repeat(1, num_images_per_prompt).view( + bs_embed * num_images_per_prompt, -1 + ) + + if self.text_encoder is not None: + if isinstance(self, StableDiffusionXLLoraLoaderMixin) and USE_PEFT_BACKEND: + # Retrieve the original scale by scaling back the LoRA layers + unscale_lora_layers(self.text_encoder, lora_scale) + + if self.text_encoder_2 is not None: + if isinstance(self, StableDiffusionXLLoraLoaderMixin) and USE_PEFT_BACKEND: + # Retrieve the original scale by scaling back the LoRA layers + unscale_lora_layers(self.text_encoder_2, lora_scale) + + return prompt_embeds, negative_prompt_embeds, pooled_prompt_embeds, negative_pooled_prompt_embeds + + # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.prepare_extra_step_kwargs + def prepare_extra_step_kwargs(self, generator, eta): + # prepare extra kwargs for the scheduler step, since not all schedulers have the same signature + # eta (η) is only used with the DDIMScheduler, it will be ignored for other schedulers. + # eta corresponds to η in DDIM paper: https://arxiv.org/abs/2010.02502 + # and should be between [0, 1] + + accepts_eta = "eta" in set(inspect.signature(self.scheduler.step).parameters.keys()) + extra_step_kwargs = {} + if accepts_eta: + extra_step_kwargs["eta"] = eta + + # check if the scheduler accepts generator + accepts_generator = "generator" in set(inspect.signature(self.scheduler.step).parameters.keys()) + if accepts_generator: + extra_step_kwargs["generator"] = generator + return extra_step_kwargs + + def check_image_size(self, x, padder_size=8): + # 获取图像的宽高 + width, height = x.size + padder_size = padder_size + # 计算需要填充的高度和宽度 + mod_pad_h = (padder_size - height % padder_size) % padder_size + mod_pad_w = (padder_size - width % padder_size) % padder_size + x_np = np.array(x) + # 使用 ImageOps.expand 进行填充 + x_padded = cv2.copyMakeBorder( + x_np, top=0, bottom=mod_pad_h, left=0, right=mod_pad_w, borderType=cv2.BORDER_REPLICATE + ) + + x = PIL.Image.fromarray(x_padded) + # x = x.resize((width + mod_pad_w, height + mod_pad_h)) + + return x, width, height, width + mod_pad_w, height + mod_pad_h + + def check_inputs( + self, + lr_img, + prompt, + prompt_2, + height, + width, + callback_steps, + negative_prompt=None, + negative_prompt_2=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 lr_img is None: + raise ValueError("`lr_image` must be provided!") + 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_steps is not None and (not isinstance(callback_steps, int) or callback_steps <= 0): + raise ValueError( + f"`callback_steps` has to be a positive integer but is {callback_steps} of type" + f" {type(callback_steps)}." + ) + + 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 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)}") + + 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." + ) + + 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`." + ) + + # Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.prepare_latents + def prepare_latents(self, batch_size, num_channels_latents, height, width, dtype, device, generator, latents=None): + shape = (batch_size, num_channels_latents, height // self.vae_scale_factor, width // self.vae_scale_factor) + 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." + ) + + if latents is None: + latents = randn_tensor(shape, generator=generator, device=device, dtype=dtype) + else: + latents = latents.to(device) + + # scale the initial noise by the standard deviation required by the scheduler + latents = latents * self.scheduler.init_noise_sigma + return latents + + def upcast_vae(self): + dtype = self.vae.dtype + self.vae.to(dtype=torch.float32) + use_torch_2_0_or_xformers = isinstance( + self.vae.decoder.mid_block.attentions[0].processor, + ( + AttnProcessor2_0, + XFormersAttnProcessor, + LoRAXFormersAttnProcessor, + LoRAAttnProcessor2_0, + FusedAttnProcessor2_0, + ), + ) + # if xformers or torch_2_0 is used attention block does not need + # to be in float32 which can save lots of memory + if use_torch_2_0_or_xformers: + self.vae.post_quant_conv.to(dtype) + self.vae.decoder.conv_in.to(dtype) + self.vae.decoder.mid_block.to(dtype) + + # Copied from diffusers.pipelines.latent_consistency_models.pipeline_latent_consistency_text2img.LatentConsistencyModelPipeline.get_guidance_scale_embedding + def get_guidance_scale_embedding( + self, w: torch.Tensor, embedding_dim: int = 512, dtype: torch.dtype = torch.float32 + ) -> torch.FloatTensor: + """ + See https://github.com/google-research/vdm/blob/dc27b98a554f65cdc654b800da5aa1846545d41b/model_vdm.py#L298 + + Args: + w (`torch.Tensor`): + Generate embedding vectors with a specified guidance scale to subsequently enrich timestep embeddings. + embedding_dim (`int`, *optional*, defaults to 512): + Dimension of the embeddings to generate. + dtype (`torch.dtype`, *optional*, defaults to `torch.float32`): + Data type of the generated embeddings. + + Returns: + `torch.FloatTensor`: Embedding vectors with shape `(len(w), embedding_dim)`. + """ + assert len(w.shape) == 1 + w = w * 1000.0 + + half_dim = embedding_dim // 2 + emb = torch.log(torch.tensor(10000.0)) / (half_dim - 1) + emb = torch.exp(torch.arange(half_dim, dtype=dtype) * -emb) + emb = w.to(dtype)[:, None] * emb[None, :] + emb = torch.cat([torch.sin(emb), torch.cos(emb)], dim=1) + if embedding_dim % 2 == 1: # zero pad + emb = torch.nn.functional.pad(emb, (0, 1)) + assert emb.shape == (w.shape[0], embedding_dim) + return emb + + def set_encoder_tile_settings( + self, + denoise_encoder_tile_sample_min_size=1024, + denoise_encoder_sample_overlap_factor=0.25, + vae_sample_size=1024, + vae_tile_overlap_factor=0.25, + ): + self.unet.denoise_encoder.tile_sample_min_size = denoise_encoder_tile_sample_min_size + self.unet.denoise_encoder.tile_overlap_factor = denoise_encoder_sample_overlap_factor + self.vae.config.sample_size = vae_sample_size + self.vae.tile_overlap_factor = vae_tile_overlap_factor + + def enable_vae_tiling(self): + r""" + Enable tiled VAE decoding. When this option is enabled, the VAE will split the input tensor into tiles to + compute decoding and encoding in several steps. This is useful for saving a large amount of memory and to allow + processing larger images. + """ + self.vae.enable_tiling() + self.unet.denoise_encoder.enable_tiling() + + def disable_vae_tiling(self): + r""" + Disable tiled VAE decoding. If `enable_vae_tiling` was previously enabled, this method will go back to + computing decoding in one step. + """ + self.vae.disable_tiling() + self.unet.denoise_encoder.disable_tiling() + + @property + def guidance_scale(self): + return self._guidance_scale + + @property + def guidance_rescale(self): + return self._guidance_rescale + + @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 and self.unet.config.time_cond_proj_dim is None + + @property + def cross_attention_kwargs(self): + return self._cross_attention_kwargs + + @property + def denoising_end(self): + return self._denoising_end + + @property + def num_timesteps(self): + return self._num_timesteps + + @property + def interrupt(self): + return self._interrupt + + 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] == 4: + image_latents = image + else: + # make sure the VAE is in float32 mode, as it overflows in float16 + # needs_upcasting = self.vae.dtype == torch.float16 and self.vae.config.force_upcast + # if needs_upcasting: + # image = image.float() + # self.upcast_vae() + self.unet.denoise_encoder.to(device=image.device, dtype=image.dtype) + image_latents = self.unet.denoise_encoder(image) + self.unet.denoise_encoder.to("cpu") + # cast back to fp16 if needed + # if needs_upcasting: + # self.vae.to(dtype=torch.float16) + + 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: + image_latents = image_latents + + if image_latents.dtype != self.vae.dtype: + image_latents = image_latents.to(dtype=self.vae.dtype) + + return image_latents + + @torch.no_grad() + @replace_example_docstring(EXAMPLE_DOC_STRING) + def __call__( + self, + lr_img: PipelineImageInput = None, + prompt: Union[str, List[str]] = None, + prompt_2: Optional[Union[str, List[str]]] = None, + height: Optional[int] = None, + width: Optional[int] = None, + num_inference_steps: int = 50, + start_point: Optional[str] = "noise", + timesteps: List[int] = None, + denoising_end: Optional[float] = None, + overlap: float = 0.5, + guidance_scale: float = 5.0, + 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, + 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, + target_size: Optional[Tuple[int, int]] = 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"], + add_sample: bool = True, + **kwargs, + ): + r""" + Function invoked when calling the pipeline for generation. + + Args: + lr_img (PipelineImageInput, optional): Low-resolution input image for conditioning the generation process. + 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. + start_point (str, *optional*): + The starting point for the generation process. Can be "noise" (random noise) or "lr" (low-resolution image). + Defaults to "noise". + 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) + overlap (float): + Overlap factor for local attention tiling (between 0.0 and 1.0). Controls the overlap between adjacent + grid patches during processing. Defaults to 0.5. + 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. + 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). + 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). + 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. + 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. + add_sample (bool): + Whether to include sample conditioning (e.g., low-resolution image) in the UNet during denoising. + Defaults to True. + + 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. + """ + + callback = kwargs.pop("callback", None) + callback_steps = kwargs.pop("callback_steps", None) + + if callback is not None: + deprecate( + "callback", + "1.0.0", + "Passing `callback` as an input argument to `__call__` is deprecated, consider use `callback_on_step_end`", + ) + if callback_steps is not None: + deprecate( + "callback_steps", + "1.0.0", + "Passing `callback_steps` as an input argument to `__call__` is deprecated, consider use `callback_on_step_end`", + ) + + # 0. Default height and width to unet + height = height or self.default_sample_size * self.vae_scale_factor + width = width or 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( + lr_img, + 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 + self._interrupt = False + self.tlc_vae_latents = LocalAttention((target_size[0] // 8, target_size[1] // 8), overlap) + self.tlc_vae_img = LocalAttention((target_size[0] // 8, target_size[1] // 8), overlap) + + # 2. Define call parameters + batch_size = 1 + num_images_per_prompt = 1 + + device = torch.device("cuda") # 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 + ) + + num_samples = num_images_per_prompt + with torch.inference_mode(): + ( + prompt_embeds, + negative_prompt_embeds, + pooled_prompt_embeds, + negative_pooled_prompt_embeds, + ) = self.encode_prompt( + prompt, + num_images_per_prompt=num_samples, + do_classifier_free_guidance=True, + negative_prompt=negative_prompt, + lora_scale=lora_scale, + ) + + lr_img_list = [lr_img] + lr_img = self.image_processor.preprocess(lr_img_list, height=height, width=width).to( + device, dtype=prompt_embeds.dtype + ) + + # 4. Prepare timesteps + timesteps, num_inference_steps = retrieve_timesteps(self.scheduler, num_inference_steps, device, timesteps) + image_latents = self.prepare_image_latents( + lr_img, batch_size, num_images_per_prompt, prompt_embeds.dtype, device, self.do_classifier_free_guidance + ) + + image_latents = self.tlc_vae_img.grids(image_latents) + + # 5. Prepare latent variables + num_channels_latents = self.vae.config.latent_channels + latents = self.prepare_latents( + batch_size * num_images_per_prompt, + num_channels_latents, + height, + width, + prompt_embeds.dtype, + device, + generator, + latents, + ) + if start_point == "lr": + latents_condition_image = self.vae.encode(lr_img * 2 - 1).latent_dist.sample() + latents_condition_image = latents_condition_image * self.vae.config.scaling_factor + start_steps_tensor = torch.randint(999, 999 + 1, (latents.shape[0],), device=latents.device) + start_steps_tensor = start_steps_tensor.long() + latents = self.DDPMScheduler.add_noise(latents_condition_image[0:1, ...], latents, start_steps_tensor) + + latents = self.tlc_vae_latents.grids(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) + views_scheduler_status = [copy.deepcopy(self.scheduler.__dict__)] * image_latents.shape[0] + + # 7. Prepare added time ids & embeddings + add_text_embeds = pooled_prompt_embeds + + 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) + + prompt_embeds = prompt_embeds.to(device) + add_text_embeds = add_text_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) + + self._num_timesteps = len(timesteps) + sub_latents_num = latents.shape[0] + with self.progress_bar(total=num_inference_steps) as progress_bar: + for i, t in enumerate(timesteps): + if i >= 1: + latents = self.tlc_vae_latents.grids(latents).to(dtype=latents.dtype) + if self.interrupt: + continue + concat_grid = [] + for sub_num in range(sub_latents_num): + self.scheduler.__dict__.update(views_scheduler_status[sub_num]) + sub_latents = latents[sub_num, :, :, :].unsqueeze(0) + img_sub_latents = image_latents[sub_num, :, :, :].unsqueeze(0) + latent_model_input = ( + torch.cat([sub_latents] * 2) if self.do_classifier_free_guidance else sub_latents + ) + img_sub_latents = ( + torch.cat([img_sub_latents] * 2) if self.do_classifier_free_guidance else img_sub_latents + ) + scaled_latent_model_input = self.scheduler.scale_model_input(latent_model_input, t) + pos_height = self.tlc_vae_latents.idxes[sub_num]["i"] + pos_width = self.tlc_vae_latents.idxes[sub_num]["j"] + add_time_ids = [ + torch.tensor([original_size]), + torch.tensor([[pos_height, pos_width]]), + torch.tensor([target_size]), + ] + add_time_ids = torch.cat(add_time_ids, dim=1).to( + img_sub_latents.device, dtype=img_sub_latents.dtype + ) + add_time_ids = add_time_ids.repeat(2, 1).to(dtype=img_sub_latents.dtype) + + # predict the noise residual + added_cond_kwargs = {"text_embeds": add_text_embeds, "time_ids": add_time_ids} + with torch.amp.autocast( + device.type, dtype=latents.dtype, enabled=latents.dtype != self.unet.dtype + ): + noise_pred = self.unet( + scaled_latent_model_input, + t, + encoder_hidden_states=prompt_embeds, + timestep_cond=timestep_cond, + cross_attention_kwargs=self.cross_attention_kwargs, + input_embedding=img_sub_latents, + add_sample=add_sample, + 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_dtype = sub_latents.dtype + sub_latents = self.scheduler.step( + noise_pred, t, sub_latents, **extra_step_kwargs, return_dict=False + )[0] + + views_scheduler_status[sub_num] = copy.deepcopy(self.scheduler.__dict__) + concat_grid.append(sub_latents) + if latents.dtype != sub_latents: + if torch.backends.mps.is_available(): + # some platforms (eg. apple mps) misbehave due to a pytorch bug: https://github.com/pytorch/pytorch/pull/99272 + sub_latents = sub_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) + 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) + + # 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) + + if XLA_AVAILABLE: + xm.mark_step() + + latents = self.tlc_vae_latents.grids_inverse(torch.cat(concat_grid, dim=0)).to(sub_latents.dtype) + + if not output_type == "latent": + # make sure the VAE is in float32 mode, as it overflows in float16 + 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) + elif latents.dtype != self.vae.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 + self.vae = self.vae.to(latents.dtype) + + # unscale/denormalize the latents + # denormalize with the mean and std if available and not None + has_latents_mean = hasattr(self.vae.config, "latents_mean") and self.vae.config.latents_mean is not None + has_latents_std = hasattr(self.vae.config, "latents_std") and self.vae.config.latents_std is not None + if has_latents_mean and has_latents_std: + latents_mean = ( + torch.tensor(self.vae.config.latents_mean).view(1, 4, 1, 1).to(latents.device, latents.dtype) + ) + latents_std = ( + torch.tensor(self.vae.config.latents_std).view(1, 4, 1, 1).to(latents.device, latents.dtype) + ) + latents = latents * latents_std / self.vae.config.scaling_factor + latents_mean + else: + latents = latents / self.vae.config.scaling_factor + + image = self.vae.decode(latents, return_dict=False)[0] + + # cast back to fp16 if needed + if needs_upcasting: + self.vae.to(dtype=torch.float16) + else: + image = latents + + if not output_type == "latent": + # apply watermark if available + if self.watermark is not None: + image = self.watermark.apply_watermark(image) + + 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 StableDiffusionXLPipelineOutput(images=image)