mirror of
https://github.com/huggingface/diffusers.git
synced 2026-01-27 17:22:53 +03:00
be4afa0bb4
* find & replace all FloatTensors to Tensor * apply formatting * Update torch.FloatTensor to torch.Tensor in the remaining files * formatting * Fix the rest of the places where FloatTensor is used as well as in documentation * formatting * Update new file from FloatTensor to Tensor
886 lines
44 KiB
Python
886 lines
44 KiB
Python
# Copyright 2024 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 inspect
|
|
from typing import Callable, List, Optional, Union
|
|
|
|
import numpy as np
|
|
import PIL.Image
|
|
import torch
|
|
from packaging import version
|
|
from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer
|
|
|
|
from diffusers import AutoencoderKL, DiffusionPipeline, UNet2DConditionModel
|
|
from diffusers.configuration_utils import FrozenDict, deprecate
|
|
from diffusers.loaders import LoraLoaderMixin, TextualInversionLoaderMixin
|
|
from diffusers.pipelines.pipeline_utils import StableDiffusionMixin
|
|
from diffusers.pipelines.stable_diffusion import StableDiffusionPipelineOutput
|
|
from diffusers.pipelines.stable_diffusion.safety_checker import (
|
|
StableDiffusionSafetyChecker,
|
|
)
|
|
from diffusers.schedulers import KarrasDiffusionSchedulers
|
|
from diffusers.utils import (
|
|
logging,
|
|
)
|
|
from diffusers.utils.torch_utils import randn_tensor
|
|
|
|
|
|
logger = logging.get_logger(__name__) # pylint: disable=invalid-name
|
|
|
|
|
|
def prepare_mask_and_masked_image(image, mask):
|
|
"""
|
|
Prepares a pair (image, mask) to be consumed by the Stable Diffusion pipeline. This means that those inputs will be
|
|
converted to ``torch.Tensor`` with shapes ``batch x channels x height x width`` where ``channels`` is ``3`` for the
|
|
``image`` and ``1`` for the ``mask``.
|
|
The ``image`` will be converted to ``torch.float32`` and normalized to be in ``[-1, 1]``. The ``mask`` will be
|
|
binarized (``mask > 0.5``) and cast to ``torch.float32`` too.
|
|
Args:
|
|
image (Union[np.array, PIL.Image, torch.Tensor]): The image to inpaint.
|
|
It can be a ``PIL.Image``, or a ``height x width x 3`` ``np.array`` or a ``channels x height x width``
|
|
``torch.Tensor`` or a ``batch x channels x height x width`` ``torch.Tensor``.
|
|
mask (_type_): The mask to apply to the image, i.e. regions to inpaint.
|
|
It can be a ``PIL.Image``, or a ``height x width`` ``np.array`` or a ``1 x height x width``
|
|
``torch.Tensor`` or a ``batch x 1 x height x width`` ``torch.Tensor``.
|
|
Raises:
|
|
ValueError: ``torch.Tensor`` images should be in the ``[-1, 1]`` range. ValueError: ``torch.Tensor`` mask
|
|
should be in the ``[0, 1]`` range. ValueError: ``mask`` and ``image`` should have the same spatial dimensions.
|
|
TypeError: ``mask`` is a ``torch.Tensor`` but ``image`` is not
|
|
(ot the other way around).
|
|
Returns:
|
|
tuple[torch.Tensor]: The pair (mask, masked_image) as ``torch.Tensor`` with 4
|
|
dimensions: ``batch x channels x height x width``.
|
|
"""
|
|
if isinstance(image, torch.Tensor):
|
|
if not isinstance(mask, torch.Tensor):
|
|
raise TypeError(f"`image` is a torch.Tensor but `mask` (type: {type(mask)} is not")
|
|
|
|
# Batch single image
|
|
if image.ndim == 3:
|
|
assert image.shape[0] == 3, "Image outside a batch should be of shape (3, H, W)"
|
|
image = image.unsqueeze(0)
|
|
|
|
# Batch and add channel dim for single mask
|
|
if mask.ndim == 2:
|
|
mask = mask.unsqueeze(0).unsqueeze(0)
|
|
|
|
# Batch single mask or add channel dim
|
|
if mask.ndim == 3:
|
|
# Single batched mask, no channel dim or single mask not batched but channel dim
|
|
if mask.shape[0] == 1:
|
|
mask = mask.unsqueeze(0)
|
|
|
|
# Batched masks no channel dim
|
|
else:
|
|
mask = mask.unsqueeze(1)
|
|
|
|
assert image.ndim == 4 and mask.ndim == 4, "Image and Mask must have 4 dimensions"
|
|
assert image.shape[-2:] == mask.shape[-2:], "Image and Mask must have the same spatial dimensions"
|
|
assert image.shape[0] == mask.shape[0], "Image and Mask must have the same batch size"
|
|
|
|
# Check image is in [-1, 1]
|
|
if image.min() < -1 or image.max() > 1:
|
|
raise ValueError("Image should be in [-1, 1] range")
|
|
|
|
# Check mask is in [0, 1]
|
|
if mask.min() < 0 or mask.max() > 1:
|
|
raise ValueError("Mask should be in [0, 1] range")
|
|
|
|
# Binarize mask
|
|
mask[mask < 0.5] = 0
|
|
mask[mask >= 0.5] = 1
|
|
|
|
# Image as float32
|
|
image = image.to(dtype=torch.float32)
|
|
elif isinstance(mask, torch.Tensor):
|
|
raise TypeError(f"`mask` is a torch.Tensor but `image` (type: {type(image)} is not")
|
|
else:
|
|
# preprocess image
|
|
if isinstance(image, (PIL.Image.Image, np.ndarray)):
|
|
image = [image]
|
|
|
|
if isinstance(image, list) and isinstance(image[0], PIL.Image.Image):
|
|
image = [np.array(i.convert("RGB"))[None, :] for i in image]
|
|
image = np.concatenate(image, axis=0)
|
|
elif isinstance(image, list) and isinstance(image[0], np.ndarray):
|
|
image = np.concatenate([i[None, :] for i in image], axis=0)
|
|
|
|
image = image.transpose(0, 3, 1, 2)
|
|
image = torch.from_numpy(image).to(dtype=torch.float32) / 127.5 - 1.0
|
|
|
|
# preprocess mask
|
|
if isinstance(mask, (PIL.Image.Image, np.ndarray)):
|
|
mask = [mask]
|
|
|
|
if isinstance(mask, list) and isinstance(mask[0], PIL.Image.Image):
|
|
mask = np.concatenate([np.array(m.convert("L"))[None, None, :] for m in mask], axis=0)
|
|
mask = mask.astype(np.float32) / 255.0
|
|
elif isinstance(mask, list) and isinstance(mask[0], np.ndarray):
|
|
mask = np.concatenate([m[None, None, :] for m in mask], axis=0)
|
|
|
|
mask[mask < 0.5] = 0
|
|
mask[mask >= 0.5] = 1
|
|
mask = torch.from_numpy(mask)
|
|
|
|
# masked_image = image * (mask >= 0.5)
|
|
masked_image = image
|
|
|
|
return mask, masked_image
|
|
|
|
|
|
class StableDiffusionRepaintPipeline(
|
|
DiffusionPipeline, StableDiffusionMixin, TextualInversionLoaderMixin, LoraLoaderMixin
|
|
):
|
|
r"""
|
|
Pipeline for text-guided image inpainting using Stable Diffusion. *This is an experimental feature*.
|
|
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.)
|
|
In addition the pipeline inherits the following loading methods:
|
|
- *Textual-Inversion*: [`loaders.TextualInversionLoaderMixin.load_textual_inversion`]
|
|
- *LoRA*: [`loaders.LoraLoaderMixin.load_lora_weights`]
|
|
as well as the following saving methods:
|
|
- *LoRA*: [`loaders.LoraLoaderMixin.save_lora_weights`]
|
|
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 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.
|
|
tokenizer (`CLIPTokenizer`):
|
|
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`].
|
|
safety_checker ([`StableDiffusionSafetyChecker`]):
|
|
Classification module that estimates whether generated images could be considered offensive or harmful.
|
|
Please, refer to the [model card](https://huggingface.co/runwayml/stable-diffusion-v1-5) for details.
|
|
feature_extractor ([`CLIPImageProcessor`]):
|
|
Model that extracts features from generated images to be used as inputs for the `safety_checker`.
|
|
"""
|
|
|
|
_optional_components = ["safety_checker", "feature_extractor"]
|
|
|
|
def __init__(
|
|
self,
|
|
vae: AutoencoderKL,
|
|
text_encoder: CLIPTextModel,
|
|
tokenizer: CLIPTokenizer,
|
|
unet: UNet2DConditionModel,
|
|
scheduler: KarrasDiffusionSchedulers,
|
|
safety_checker: StableDiffusionSafetyChecker,
|
|
feature_extractor: CLIPImageProcessor,
|
|
requires_safety_checker: bool = True,
|
|
):
|
|
super().__init__()
|
|
|
|
if hasattr(scheduler.config, "steps_offset") and scheduler.config.steps_offset != 1:
|
|
deprecation_message = (
|
|
f"The configuration file of this scheduler: {scheduler} is outdated. `steps_offset`"
|
|
f" should be set to 1 instead of {scheduler.config.steps_offset}. Please make sure "
|
|
"to update the config accordingly as leaving `steps_offset` might led to incorrect results"
|
|
" in future versions. If you have downloaded this checkpoint from the Hugging Face Hub,"
|
|
" it would be very nice if you could open a Pull request for the `scheduler/scheduler_config.json`"
|
|
" file"
|
|
)
|
|
deprecate("steps_offset!=1", "1.0.0", deprecation_message, standard_warn=False)
|
|
new_config = dict(scheduler.config)
|
|
new_config["steps_offset"] = 1
|
|
scheduler._internal_dict = FrozenDict(new_config)
|
|
|
|
if hasattr(scheduler.config, "skip_prk_steps") and scheduler.config.skip_prk_steps is False:
|
|
deprecation_message = (
|
|
f"The configuration file of this scheduler: {scheduler} has not set the configuration"
|
|
" `skip_prk_steps`. `skip_prk_steps` should be set to True in the configuration file. Please make"
|
|
" sure to update the config accordingly as not setting `skip_prk_steps` in the config might lead to"
|
|
" incorrect results in future versions. If you have downloaded this checkpoint from the Hugging Face"
|
|
" Hub, it would be very nice if you could open a Pull request for the"
|
|
" `scheduler/scheduler_config.json` file"
|
|
)
|
|
deprecate(
|
|
"skip_prk_steps not set",
|
|
"1.0.0",
|
|
deprecation_message,
|
|
standard_warn=False,
|
|
)
|
|
new_config = dict(scheduler.config)
|
|
new_config["skip_prk_steps"] = True
|
|
scheduler._internal_dict = FrozenDict(new_config)
|
|
|
|
if safety_checker is None and requires_safety_checker:
|
|
logger.warning(
|
|
f"You have disabled the safety checker for {self.__class__} by passing `safety_checker=None`. Ensure"
|
|
" that you abide to the conditions of the Stable Diffusion license and do not expose unfiltered"
|
|
" results in services or applications open to the public. Both the diffusers team and Hugging Face"
|
|
" strongly recommend to keep the safety filter enabled in all public facing circumstances, disabling"
|
|
" it only for use-cases that involve analyzing network behavior or auditing its results. For more"
|
|
" information, please have a look at https://github.com/huggingface/diffusers/pull/254 ."
|
|
)
|
|
|
|
if safety_checker is not None and feature_extractor is None:
|
|
raise ValueError(
|
|
"Make sure to define a feature extractor when loading {self.__class__} if you want to use the safety"
|
|
" checker. If you do not want to use the safety checker, you can pass `'safety_checker=None'` instead."
|
|
)
|
|
|
|
is_unet_version_less_0_9_0 = hasattr(unet.config, "_diffusers_version") and version.parse(
|
|
version.parse(unet.config._diffusers_version).base_version
|
|
) < version.parse("0.9.0.dev0")
|
|
is_unet_sample_size_less_64 = hasattr(unet.config, "sample_size") and unet.config.sample_size < 64
|
|
if is_unet_version_less_0_9_0 and is_unet_sample_size_less_64:
|
|
deprecation_message = (
|
|
"The configuration file of the unet has set the default `sample_size` to smaller than"
|
|
" 64 which seems highly unlikely .If you're checkpoint is a fine-tuned version of any of the"
|
|
" following: \n- CompVis/stable-diffusion-v1-4 \n- CompVis/stable-diffusion-v1-3 \n-"
|
|
" CompVis/stable-diffusion-v1-2 \n- CompVis/stable-diffusion-v1-1 \n- runwayml/stable-diffusion-v1-5"
|
|
" \n- runwayml/stable-diffusion-inpainting \n you should change 'sample_size' to 64 in the"
|
|
" configuration file. Please make sure to update the config accordingly as leaving `sample_size=32`"
|
|
" in the config might lead to incorrect results in future versions. If you have downloaded this"
|
|
" checkpoint from the Hugging Face Hub, it would be very nice if you could open a Pull request for"
|
|
" the `unet/config.json` file"
|
|
)
|
|
deprecate("sample_size<64", "1.0.0", deprecation_message, standard_warn=False)
|
|
new_config = dict(unet.config)
|
|
new_config["sample_size"] = 64
|
|
unet._internal_dict = FrozenDict(new_config)
|
|
# Check shapes, assume num_channels_latents == 4, num_channels_mask == 1, num_channels_masked == 4
|
|
if unet.config.in_channels != 4:
|
|
logger.warning(
|
|
f"You have loaded a UNet with {unet.config.in_channels} input channels, whereas by default,"
|
|
f" {self.__class__} assumes that `pipeline.unet` has 4 input channels: 4 for `num_channels_latents`,"
|
|
". If you did not intend to modify"
|
|
" this behavior, please check whether you have loaded the right checkpoint."
|
|
)
|
|
|
|
self.register_modules(
|
|
vae=vae,
|
|
text_encoder=text_encoder,
|
|
tokenizer=tokenizer,
|
|
unet=unet,
|
|
scheduler=scheduler,
|
|
safety_checker=safety_checker,
|
|
feature_extractor=feature_extractor,
|
|
)
|
|
self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
|
|
self.register_to_config(requires_safety_checker=requires_safety_checker)
|
|
|
|
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline._encode_prompt
|
|
def _encode_prompt(
|
|
self,
|
|
prompt,
|
|
device,
|
|
num_images_per_prompt,
|
|
do_classifier_free_guidance,
|
|
negative_prompt=None,
|
|
prompt_embeds: Optional[torch.Tensor] = None,
|
|
negative_prompt_embeds: Optional[torch.Tensor] = None,
|
|
):
|
|
r"""
|
|
Encodes the prompt into text encoder hidden states.
|
|
Args:
|
|
prompt (`str` or `List[str]`, *optional*):
|
|
prompt to be encoded
|
|
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`).
|
|
prompt_embeds (`torch.Tensor`, *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.Tensor`, *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.
|
|
"""
|
|
if prompt is not None and isinstance(prompt, str):
|
|
batch_size = 1
|
|
elif prompt is not None and isinstance(prompt, list):
|
|
batch_size = len(prompt)
|
|
else:
|
|
batch_size = prompt_embeds.shape[0]
|
|
|
|
if prompt_embeds is None:
|
|
# textual inversion: process multi-vector tokens if necessary
|
|
if isinstance(self, TextualInversionLoaderMixin):
|
|
prompt = self.maybe_convert_prompt(prompt, self.tokenizer)
|
|
|
|
text_inputs = self.tokenizer(
|
|
prompt,
|
|
padding="max_length",
|
|
max_length=self.tokenizer.model_max_length,
|
|
truncation=True,
|
|
return_tensors="pt",
|
|
)
|
|
text_input_ids = text_inputs.input_ids
|
|
untruncated_ids = self.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 = self.tokenizer.batch_decode(
|
|
untruncated_ids[:, self.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" {self.tokenizer.model_max_length} tokens: {removed_text}"
|
|
)
|
|
|
|
if hasattr(self.text_encoder.config, "use_attention_mask") and self.text_encoder.config.use_attention_mask:
|
|
attention_mask = text_inputs.attention_mask.to(device)
|
|
else:
|
|
attention_mask = None
|
|
|
|
prompt_embeds = self.text_encoder(
|
|
text_input_ids.to(device),
|
|
attention_mask=attention_mask,
|
|
)
|
|
prompt_embeds = prompt_embeds[0]
|
|
|
|
prompt_embeds = prompt_embeds.to(dtype=self.text_encoder.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)
|
|
|
|
# get unconditional embeddings for classifier free guidance
|
|
if do_classifier_free_guidance and negative_prompt_embeds is None:
|
|
uncond_tokens: List[str]
|
|
if negative_prompt is None:
|
|
uncond_tokens = [""] * batch_size
|
|
elif 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 isinstance(negative_prompt, str):
|
|
uncond_tokens = [negative_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
|
|
|
|
# textual inversion: process multi-vector tokens if necessary
|
|
if isinstance(self, TextualInversionLoaderMixin):
|
|
uncond_tokens = self.maybe_convert_prompt(uncond_tokens, self.tokenizer)
|
|
|
|
max_length = prompt_embeds.shape[1]
|
|
uncond_input = self.tokenizer(
|
|
uncond_tokens,
|
|
padding="max_length",
|
|
max_length=max_length,
|
|
truncation=True,
|
|
return_tensors="pt",
|
|
)
|
|
|
|
if hasattr(self.text_encoder.config, "use_attention_mask") and self.text_encoder.config.use_attention_mask:
|
|
attention_mask = uncond_input.attention_mask.to(device)
|
|
else:
|
|
attention_mask = None
|
|
|
|
negative_prompt_embeds = self.text_encoder(
|
|
uncond_input.input_ids.to(device),
|
|
attention_mask=attention_mask,
|
|
)
|
|
negative_prompt_embeds = negative_prompt_embeds[0]
|
|
|
|
if do_classifier_free_guidance:
|
|
# duplicate unconditional embeddings for each generation per prompt, using mps friendly method
|
|
seq_len = negative_prompt_embeds.shape[1]
|
|
|
|
negative_prompt_embeds = negative_prompt_embeds.to(dtype=self.text_encoder.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)
|
|
|
|
# For classifier free guidance, we need to do two forward passes.
|
|
# Here we concatenate the unconditional and text embeddings into a single batch
|
|
# to avoid doing two forward passes
|
|
prompt_embeds = torch.cat([negative_prompt_embeds, prompt_embeds])
|
|
|
|
return prompt_embeds
|
|
|
|
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.run_safety_checker
|
|
def run_safety_checker(self, image, device, dtype):
|
|
if self.safety_checker is not None:
|
|
safety_checker_input = self.feature_extractor(self.numpy_to_pil(image), return_tensors="pt").to(device)
|
|
image, has_nsfw_concept = self.safety_checker(
|
|
images=image, clip_input=safety_checker_input.pixel_values.to(dtype)
|
|
)
|
|
else:
|
|
has_nsfw_concept = None
|
|
return image, has_nsfw_concept
|
|
|
|
# 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
|
|
|
|
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.decode_latents
|
|
def decode_latents(self, latents):
|
|
latents = 1 / self.vae.config.scaling_factor * latents
|
|
image = self.vae.decode(latents).sample
|
|
image = (image / 2 + 0.5).clamp(0, 1)
|
|
# we always cast to float32 as this does not cause significant overhead and is compatible with bfloat16
|
|
image = image.cpu().permute(0, 2, 3, 1).float().numpy()
|
|
return image
|
|
|
|
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.check_inputs
|
|
def check_inputs(
|
|
self,
|
|
prompt,
|
|
height,
|
|
width,
|
|
callback_steps,
|
|
negative_prompt=None,
|
|
prompt_embeds=None,
|
|
negative_prompt_embeds=None,
|
|
):
|
|
if height % 8 != 0 or width % 8 != 0:
|
|
raise ValueError(f"`height` and `width` have to be divisible by 8 but are {height} and {width}.")
|
|
|
|
if (callback_steps is None) or (
|
|
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 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 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)}")
|
|
|
|
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."
|
|
)
|
|
|
|
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}."
|
|
)
|
|
|
|
# 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 prepare_mask_latents(
|
|
self,
|
|
mask,
|
|
masked_image,
|
|
batch_size,
|
|
height,
|
|
width,
|
|
dtype,
|
|
device,
|
|
generator,
|
|
do_classifier_free_guidance,
|
|
):
|
|
# resize the mask to latents shape as we concatenate the mask to the latents
|
|
# we do that before converting to dtype to avoid breaking in case we're using cpu_offload
|
|
# and half precision
|
|
mask = torch.nn.functional.interpolate(
|
|
mask, size=(height // self.vae_scale_factor, width // self.vae_scale_factor)
|
|
)
|
|
mask = mask.to(device=device, dtype=dtype)
|
|
|
|
masked_image = masked_image.to(device=device, dtype=dtype)
|
|
|
|
# encode the mask image into latents space so we can concatenate it to the latents
|
|
if isinstance(generator, list):
|
|
masked_image_latents = [
|
|
self.vae.encode(masked_image[i : i + 1]).latent_dist.sample(generator=generator[i])
|
|
for i in range(batch_size)
|
|
]
|
|
masked_image_latents = torch.cat(masked_image_latents, dim=0)
|
|
else:
|
|
masked_image_latents = self.vae.encode(masked_image).latent_dist.sample(generator=generator)
|
|
masked_image_latents = self.vae.config.scaling_factor * masked_image_latents
|
|
|
|
# duplicate mask and masked_image_latents for each generation per prompt, using mps friendly method
|
|
if mask.shape[0] < batch_size:
|
|
if not batch_size % mask.shape[0] == 0:
|
|
raise ValueError(
|
|
"The passed mask and the required batch size don't match. Masks are supposed to be duplicated to"
|
|
f" a total batch size of {batch_size}, but {mask.shape[0]} masks were passed. Make sure the number"
|
|
" of masks that you pass is divisible by the total requested batch size."
|
|
)
|
|
mask = mask.repeat(batch_size // mask.shape[0], 1, 1, 1)
|
|
if masked_image_latents.shape[0] < batch_size:
|
|
if not batch_size % masked_image_latents.shape[0] == 0:
|
|
raise ValueError(
|
|
"The passed images and the required batch size don't match. Images are supposed to be duplicated"
|
|
f" to a total batch size of {batch_size}, but {masked_image_latents.shape[0]} images were passed."
|
|
" Make sure the number of images that you pass is divisible by the total requested batch size."
|
|
)
|
|
masked_image_latents = masked_image_latents.repeat(batch_size // masked_image_latents.shape[0], 1, 1, 1)
|
|
|
|
mask = torch.cat([mask] * 2) if do_classifier_free_guidance else mask
|
|
masked_image_latents = (
|
|
torch.cat([masked_image_latents] * 2) if do_classifier_free_guidance else masked_image_latents
|
|
)
|
|
|
|
# aligning device to prevent device errors when concating it with the latent model input
|
|
masked_image_latents = masked_image_latents.to(device=device, dtype=dtype)
|
|
return mask, masked_image_latents
|
|
|
|
@torch.no_grad()
|
|
def __call__(
|
|
self,
|
|
prompt: Union[str, List[str]] = None,
|
|
image: Union[torch.Tensor, PIL.Image.Image] = None,
|
|
mask_image: Union[torch.Tensor, PIL.Image.Image] = None,
|
|
height: Optional[int] = None,
|
|
width: Optional[int] = None,
|
|
num_inference_steps: int = 50,
|
|
jump_length: Optional[int] = 10,
|
|
jump_n_sample: Optional[int] = 10,
|
|
guidance_scale: float = 7.5,
|
|
negative_prompt: 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.Tensor] = None,
|
|
prompt_embeds: Optional[torch.Tensor] = None,
|
|
negative_prompt_embeds: Optional[torch.Tensor] = None,
|
|
output_type: Optional[str] = "pil",
|
|
return_dict: bool = True,
|
|
callback: Optional[Callable[[int, int, torch.Tensor], None]] = None,
|
|
callback_steps: int = 1,
|
|
):
|
|
r"""
|
|
Function invoked when calling the pipeline for generation.
|
|
Args:
|
|
prompt (`str` or `List[str]`, *optional*):
|
|
The prompt or prompts to guide the image generation. If not defined, one has to pass `prompt_embeds`.
|
|
instead.
|
|
image (`PIL.Image.Image`):
|
|
`Image`, or tensor representing an image batch which will be inpainted, *i.e.* parts of the image will
|
|
be masked out with `mask_image` and repainted according to `prompt`.
|
|
mask_image (`PIL.Image.Image`):
|
|
`Image`, or tensor representing an image batch, to mask `image`. White pixels in the mask will be
|
|
repainted, while black pixels will be preserved. If `mask_image` is a PIL image, it will be converted
|
|
to a single channel (luminance) before use. If it's a tensor, it should contain one color channel (L)
|
|
instead of 3, so the expected shape would be `(B, H, W, 1)`.
|
|
height (`int`, *optional*, defaults to self.unet.config.sample_size * self.vae_scale_factor):
|
|
The height in pixels of the generated image.
|
|
width (`int`, *optional*, defaults to self.unet.config.sample_size * self.vae_scale_factor):
|
|
The width in pixels of the generated image.
|
|
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.
|
|
jump_length (`int`, *optional*, defaults to 10):
|
|
The number of steps taken forward in time before going backward in time for a single jump ("j" in
|
|
RePaint paper). Take a look at Figure 9 and 10 in https://arxiv.org/pdf/2201.09865.pdf.
|
|
jump_n_sample (`int`, *optional*, defaults to 10):
|
|
The number of times we will make forward time jump for a given chosen time sample. Take a look at
|
|
Figure 9 and 10 in https://arxiv.org/pdf/2201.09865.pdf.
|
|
guidance_scale (`float`, *optional*, defaults to 7.5):
|
|
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`).
|
|
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`, *optional*):
|
|
One or a list of [torch generator(s)](https://pytorch.org/docs/stable/generated/torch.Generator.html)
|
|
to make generation deterministic.
|
|
latents (`torch.Tensor`, *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.Tensor`, *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.Tensor`, *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.
|
|
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.StableDiffusionPipelineOutput`] instead of a
|
|
plain tuple.
|
|
callback (`Callable`, *optional*):
|
|
A function that will be called every `callback_steps` steps during inference. The function will be
|
|
called with the following arguments: `callback(step: int, timestep: int, latents: torch.Tensor)`.
|
|
callback_steps (`int`, *optional*, defaults to 1):
|
|
The frequency at which the `callback` function will be called. If not specified, the callback will be
|
|
called at every step.
|
|
Examples:
|
|
```py
|
|
>>> import PIL
|
|
>>> import requests
|
|
>>> import torch
|
|
>>> from io import BytesIO
|
|
>>> from diffusers import StableDiffusionPipeline, RePaintScheduler
|
|
>>> def download_image(url):
|
|
... response = requests.get(url)
|
|
... return PIL.Image.open(BytesIO(response.content)).convert("RGB")
|
|
>>> base_url = "https://raw.githubusercontent.com/CompVis/latent-diffusion/main/data/inpainting_examples/"
|
|
>>> img_url = base_url + "overture-creations-5sI6fQgYIuo.png"
|
|
>>> mask_url = base_url + "overture-creations-5sI6fQgYIuo_mask.png "
|
|
>>> init_image = download_image(img_url).resize((512, 512))
|
|
>>> mask_image = download_image(mask_url).resize((512, 512))
|
|
>>> pipe = DiffusionPipeline.from_pretrained(
|
|
... "CompVis/stable-diffusion-v1-4", torch_dtype=torch.float16, custom_pipeline="stable_diffusion_repaint",
|
|
... )
|
|
>>> pipe.scheduler = RePaintScheduler.from_config(pipe.scheduler.config)
|
|
>>> pipe = pipe.to("cuda")
|
|
>>> prompt = "Face of a yellow cat, high resolution, sitting on a park bench"
|
|
>>> image = pipe(prompt=prompt, image=init_image, mask_image=mask_image).images[0]
|
|
```
|
|
Returns:
|
|
[`~pipelines.stable_diffusion.StableDiffusionPipelineOutput`] or `tuple`:
|
|
[`~pipelines.stable_diffusion.StableDiffusionPipelineOutput`] if `return_dict` is True, otherwise a `tuple.
|
|
When returning a tuple, the first element is a list with the generated images, and the second element is a
|
|
list of `bool`s denoting whether the corresponding generated image likely represents "not-safe-for-work"
|
|
(nsfw) content, according to the `safety_checker`.
|
|
"""
|
|
# 0. Default height and width to unet
|
|
height = height or self.unet.config.sample_size * self.vae_scale_factor
|
|
width = width or self.unet.config.sample_size * self.vae_scale_factor
|
|
|
|
# 1. Check inputs
|
|
self.check_inputs(
|
|
prompt,
|
|
height,
|
|
width,
|
|
callback_steps,
|
|
negative_prompt,
|
|
prompt_embeds,
|
|
negative_prompt_embeds,
|
|
)
|
|
|
|
if image is None:
|
|
raise ValueError("`image` input cannot be undefined.")
|
|
|
|
if mask_image is None:
|
|
raise ValueError("`mask_image` input cannot be undefined.")
|
|
|
|
# 2. Define call parameters
|
|
if prompt is not None and isinstance(prompt, str):
|
|
batch_size = 1
|
|
elif prompt is not None and isinstance(prompt, list):
|
|
batch_size = len(prompt)
|
|
else:
|
|
batch_size = prompt_embeds.shape[0]
|
|
|
|
device = self._execution_device
|
|
# 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.
|
|
do_classifier_free_guidance = guidance_scale > 1.0
|
|
|
|
# 3. Encode input prompt
|
|
prompt_embeds = self._encode_prompt(
|
|
prompt,
|
|
device,
|
|
num_images_per_prompt,
|
|
do_classifier_free_guidance,
|
|
negative_prompt,
|
|
prompt_embeds=prompt_embeds,
|
|
negative_prompt_embeds=negative_prompt_embeds,
|
|
)
|
|
|
|
# 4. Preprocess mask and image
|
|
mask, masked_image = prepare_mask_and_masked_image(image, mask_image)
|
|
|
|
# 5. set timesteps
|
|
self.scheduler.set_timesteps(num_inference_steps, jump_length, jump_n_sample, device)
|
|
self.scheduler.eta = eta
|
|
|
|
timesteps = self.scheduler.timesteps
|
|
# latent_timestep = timesteps[:1].repeat(batch_size * num_images_per_prompt)
|
|
|
|
# 6. 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,
|
|
)
|
|
|
|
# 7. Prepare mask latent variables
|
|
mask, masked_image_latents = self.prepare_mask_latents(
|
|
mask,
|
|
masked_image,
|
|
batch_size * num_images_per_prompt,
|
|
height,
|
|
width,
|
|
prompt_embeds.dtype,
|
|
device,
|
|
generator,
|
|
do_classifier_free_guidance=False, # We do not need duplicate mask and image
|
|
)
|
|
|
|
# 8. Check that sizes of mask, masked image and latents match
|
|
# num_channels_mask = mask.shape[1]
|
|
# num_channels_masked_image = masked_image_latents.shape[1]
|
|
if num_channels_latents != self.unet.config.in_channels:
|
|
raise ValueError(
|
|
f"Incorrect configuration settings! The config of `pipeline.unet`: {self.unet.config} expects"
|
|
f" {self.unet.config.in_channels} but received `num_channels_latents`: {num_channels_latents} "
|
|
f" = Please verify the config of"
|
|
" `pipeline.unet` or your `mask_image` or `image` input."
|
|
)
|
|
|
|
# 9. 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)
|
|
|
|
t_last = timesteps[0] + 1
|
|
|
|
# 10. Denoising loop
|
|
with self.progress_bar(total=len(timesteps)) as progress_bar:
|
|
for i, t in enumerate(timesteps):
|
|
if t >= t_last:
|
|
# compute the reverse: x_t-1 -> x_t
|
|
latents = self.scheduler.undo_step(latents, t_last, generator)
|
|
progress_bar.update()
|
|
t_last = t
|
|
continue
|
|
|
|
# expand the latents if we are doing classifier free guidance
|
|
latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents
|
|
|
|
# concat latents, mask, masked_image_latents in the channel dimension
|
|
latent_model_input = self.scheduler.scale_model_input(latent_model_input, t)
|
|
# latent_model_input = torch.cat([latent_model_input, mask, masked_image_latents], dim=1)
|
|
|
|
# predict the noise residual
|
|
noise_pred = self.unet(latent_model_input, t, encoder_hidden_states=prompt_embeds).sample
|
|
|
|
# perform guidance
|
|
if do_classifier_free_guidance:
|
|
noise_pred_uncond, noise_pred_text = noise_pred.chunk(2)
|
|
noise_pred = noise_pred_uncond + guidance_scale * (noise_pred_text - noise_pred_uncond)
|
|
|
|
# compute the previous noisy sample x_t -> x_t-1
|
|
latents = self.scheduler.step(
|
|
noise_pred,
|
|
t,
|
|
latents,
|
|
masked_image_latents,
|
|
mask,
|
|
**extra_step_kwargs,
|
|
).prev_sample
|
|
|
|
# call the callback, if provided
|
|
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)
|
|
|
|
t_last = t
|
|
|
|
# 11. Post-processing
|
|
image = self.decode_latents(latents)
|
|
|
|
# 12. Run safety checker
|
|
image, has_nsfw_concept = self.run_safety_checker(image, device, prompt_embeds.dtype)
|
|
|
|
# 13. Convert to PIL
|
|
if output_type == "pil":
|
|
image = self.numpy_to_pil(image)
|
|
|
|
# Offload last model to CPU
|
|
if hasattr(self, "final_offload_hook") and self.final_offload_hook is not None:
|
|
self.final_offload_hook.offload()
|
|
|
|
if not return_dict:
|
|
return (image, has_nsfw_concept)
|
|
|
|
return StableDiffusionPipelineOutput(images=image, nsfw_content_detected=has_nsfw_concept)
|