AI/diffusers
1
0
Fork 0
mirror of https://github.com/huggingface/diffusers.git synced 2026-01-29 07:22:12 +03:00
Code Activity

Fixing implementation of ControlNet-XS (#6772)

Browse Source 
* CheckIn - created DownSubBlocks

* Added extra channels, implemented subblock fwd

* Fixed connection sizes

* checkin

* Removed iter, next in forward

* Models for SD21 & SDXL run through

* Added back pipelines, cleared up connections

* Cleaned up connection creation

* added debug logs

* updated logs

* logs: added input loading

* Update umer_debug_logger.py

* log: Loading hint

* Update umer_debug_logger.py

* added logs

* Changed debug logging

* debug: added more logs

* Fixed num_norm_groups

* Debug: Logging all of SDXL input

* Update umer_debug_logger.py

* debug: updated logs

* checkim

* Readded tests

* Removed debug logs

* Fixed Slow Tests

* Added value ckecks | Updated model_cpu_offload_seq

* accelerate-offloading works ; fast tests work

* Made unet & addon explicit in controlnet

* Updated slow tests

* Added dtype/device to ControlNetXS

* Filled in test model paths

* Added image_encoder/feature_extractor to XL pipe

* Fixed fast tests

* Added comments and docstrings

* Fixed copies

* Added docs ; Updates slow tests

* Moved changes to UNetMidBlock2DCrossAttn

* tiny cleanups

* Removed stray prints

* Removed ip adapters + freeU

- Removed ip adapters + freeU as they don't make sense for ControlNet-XS
- Fixed imports of UNet components

* Fixed test_save_load_float16

* Make style, quality, fix-copies

* Changed loading/saving API for ControlNetXS

- Changed loading/saving API for ControlNetXS
- other small fixes

* Removed ControlNet-XS from research examples

* Make style, quality, fix-copies

* Small fixes

- deleted ControlNetXSModel.init_original
- added time_embedding_mix to StableDiffusionControlNetXSPipeline .from_pretrained / StableDiffusionXLControlNetXSPipeline.from_pretrained
- fixed copy hints

* checkin May 11 '23

* CheckIn Mar 12 '24

* Fixed tests for SD

* Added tests for UNetControlNetXSModel

* Fixed SDXL tests

* cleanup

* Delete Pipfile

* CheckIn Mar 20

Started replacing sub blocks  by `ControlNetXSCrossAttnDownBlock2D` and `ControlNetXSCrossAttnUplock2D`

* check-in Mar 23

* checkin 24 Mar

* Created init for UNetCnxs and CnxsAddon

* CheckIn

* Made from_modules, from_unet and no_control work

* make style,quality,fix-copies & small changes

* Fixed freezing

* Added gradient ckpt'ing; fixed tests

* Fix slow tests(+compile) ; clear naming confusion

* Don't create UNet in init ; removed class_emb

* Incorporated review feedback

- Deleted get_base_pipeline /  get_controlnet_addon for pipes
- Pipes inherit from StableDiffusionXLPipeline
- Made module dicts for cnxs-addon's down/mid/up classes
- Added support for qkv fusion and freeU

* Make style, quality, fix-copies

* Implemented review feedback

* Removed compatibility check for vae/ctrl embedding

* make style, quality, fix-copies

* Delete Pipfile

* Integrated review feedback

- Importing ControlNetConditioningEmbedding now
- get_down/mid/up_block_addon now outside class
- renamed `do_control` to `apply_control`

* Reduced size of test tensors

For this, added `norm_num_groups` as parameter everywhere

* Renamed cnxs-`Addon` to cnxs-`Adapter`

- `ControlNetXSAddon` -> `ControlNetXSAdapter`
- `ControlNetXSAddonDownBlockComponents` -> `DownBlockControlNetXSAdapter`, and similarly for mid/up
- `get_mid_block_addon` -> `get_mid_block_adapter`, and similarly for mid/up

* Fixed save_pretrained/from_pretrained bug

* Removed redundant code

---------

Co-authored-by: Dhruv Nair <dhruv.nair@gmail.com>
This commit is contained in:
UmerHA
2024-04-16 18:26:20 +02:00
committed by GitHub
parent cf6e0407e0
commit fda1531d8a
22 changed files with 3640 additions and 1338 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
src/diffusers/__init__.py
View File

@@ -80,6 +80,7 @@ else:
"AutoencoderTiny",
"ConsistencyDecoderVAE",
"ControlNetModel",
"ControlNetXSAdapter",
"I2VGenXLUNet",
"Kandinsky3UNet",
"ModelMixin",
@@ -94,6 +95,7 @@ else:
"UNet2DConditionModel",
"UNet2DModel",
"UNet3DConditionModel",
"UNetControlNetXSModel",
"UNetMotionModel",
"UNetSpatioTemporalConditionModel",
"UVit2DModel",
@@ -270,6 +272,7 @@ else:
"StableDiffusionControlNetImg2ImgPipeline",
"StableDiffusionControlNetInpaintPipeline",
"StableDiffusionControlNetPipeline",
"StableDiffusionControlNetXSPipeline",
"StableDiffusionDepth2ImgPipeline",
"StableDiffusionDiffEditPipeline",
"StableDiffusionGLIGENPipeline",
@@ -293,6 +296,7 @@ else:
"StableDiffusionXLControlNetImg2ImgPipeline",
"StableDiffusionXLControlNetInpaintPipeline",
"StableDiffusionXLControlNetPipeline",
"StableDiffusionXLControlNetXSPipeline",
"StableDiffusionXLImg2ImgPipeline",
"StableDiffusionXLInpaintPipeline",
"StableDiffusionXLInstructPix2PixPipeline",
@@ -474,6 +478,7 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
AutoencoderTiny,
ConsistencyDecoderVAE,
ControlNetModel,
ControlNetXSAdapter,
I2VGenXLUNet,
Kandinsky3UNet,
ModelMixin,
@@ -487,6 +492,7 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
UNet2DConditionModel,
UNet2DModel,
UNet3DConditionModel,
UNetControlNetXSModel,
UNetMotionModel,
UNetSpatioTemporalConditionModel,
UVit2DModel,
@@ -642,6 +648,7 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
StableDiffusionControlNetImg2ImgPipeline,
StableDiffusionControlNetInpaintPipeline,
StableDiffusionControlNetPipeline,
StableDiffusionControlNetXSPipeline,
StableDiffusionDepth2ImgPipeline,
StableDiffusionDiffEditPipeline,
StableDiffusionGLIGENPipeline,
@@ -665,6 +672,7 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
StableDiffusionXLControlNetImg2ImgPipeline,
StableDiffusionXLControlNetInpaintPipeline,
StableDiffusionXLControlNetPipeline,
StableDiffusionXLControlNetXSPipeline,
StableDiffusionXLImg2ImgPipeline,
StableDiffusionXLInpaintPipeline,
StableDiffusionXLInstructPix2PixPipeline,

2
src/diffusers/models/__init__.py
View File

@@ -32,6 +32,7 @@ if is_torch_available():
_import_structure["autoencoders.autoencoder_tiny"] = ["AutoencoderTiny"]
_import_structure["autoencoders.consistency_decoder_vae"] = ["ConsistencyDecoderVAE"]
_import_structure["controlnet"] = ["ControlNetModel"]
_import_structure["controlnet_xs"] = ["ControlNetXSAdapter", "UNetControlNetXSModel"]
_import_structure["dual_transformer_2d"] = ["DualTransformer2DModel"]
_import_structure["embeddings"] = ["ImageProjection"]
_import_structure["modeling_utils"] = ["ModelMixin"]
@@ -68,6 +69,7 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
ConsistencyDecoderVAE,
)
from .controlnet import ControlNetModel
from .controlnet_xs import ControlNetXSAdapter, UNetControlNetXSModel
from .embeddings import ImageProjection
from .modeling_utils import ModelMixin
from .transformers import (

1892
src/diffusers/models/controlnet_xs.py Normal file
View File

File diff suppressed because it is too large Load Diff

27
src/diffusers/models/unets/unet_2d_blocks.py
View File

@@ -746,6 +746,7 @@ class UNetMidBlock2DCrossAttn(nn.Module):
self,
in_channels: int,
temb_channels: int,
out_channels: Optional[int] = None,
dropout: float = 0.0,
num_layers: int = 1,
transformer_layers_per_block: Union[int, Tuple[int]] = 1,
@@ -753,6 +754,7 @@ class UNetMidBlock2DCrossAttn(nn.Module):
resnet_time_scale_shift: str = "default",
resnet_act_fn: str = "swish",
resnet_groups: int = 32,
resnet_groups_out: Optional[int] = None,
resnet_pre_norm: bool = True,
num_attention_heads: int = 1,
output_scale_factor: float = 1.0,
@@ -764,6 +766,10 @@ class UNetMidBlock2DCrossAttn(nn.Module):
):
super().__init__()
out_channels = out_channels or in_channels
self.in_channels = in_channels
self.out_channels = out_channels
self.has_cross_attention = True
self.num_attention_heads = num_attention_heads
resnet_groups = resnet_groups if resnet_groups is not None else min(in_channels // 4, 32)
@@ -772,14 +778,17 @@ class UNetMidBlock2DCrossAttn(nn.Module):
if isinstance(transformer_layers_per_block, int):
transformer_layers_per_block = [transformer_layers_per_block] * num_layers
resnet_groups_out = resnet_groups_out or resnet_groups
# there is always at least one resnet
resnets = [
ResnetBlock2D(
in_channels=in_channels,
out_channels=in_channels,
out_channels=out_channels,
temb_channels=temb_channels,
eps=resnet_eps,
groups=resnet_groups,
groups_out=resnet_groups_out,
dropout=dropout,
time_embedding_norm=resnet_time_scale_shift,
non_linearity=resnet_act_fn,
@@ -794,11 +803,11 @@ class UNetMidBlock2DCrossAttn(nn.Module):
attentions.append(
Transformer2DModel(
num_attention_heads,
in_channels // num_attention_heads,
in_channels=in_channels,
out_channels // num_attention_heads,
in_channels=out_channels,
num_layers=transformer_layers_per_block[i],
cross_attention_dim=cross_attention_dim,
norm_num_groups=resnet_groups,
norm_num_groups=resnet_groups_out,
use_linear_projection=use_linear_projection,
upcast_attention=upcast_attention,
attention_type=attention_type,
@@ -808,8 +817,8 @@ class UNetMidBlock2DCrossAttn(nn.Module):
attentions.append(
DualTransformer2DModel(
num_attention_heads,
in_channels // num_attention_heads,
in_channels=in_channels,
out_channels // num_attention_heads,
in_channels=out_channels,
num_layers=1,
cross_attention_dim=cross_attention_dim,
norm_num_groups=resnet_groups,
@@ -817,11 +826,11 @@ class UNetMidBlock2DCrossAttn(nn.Module):
)
resnets.append(
ResnetBlock2D(
in_channels=in_channels,
out_channels=in_channels,
in_channels=out_channels,
out_channels=out_channels,
temb_channels=temb_channels,
eps=resnet_eps,
groups=resnet_groups,
groups=resnet_groups_out,
dropout=dropout,
time_embedding_norm=resnet_time_scale_shift,
non_linearity=resnet_act_fn,

10
src/diffusers/pipelines/__init__.py
View File

@@ -134,6 +134,12 @@ else:
"StableDiffusionXLControlNetPipeline",
]
)
_import_structure["controlnet_xs"].extend(
[
"StableDiffusionControlNetXSPipeline",
"StableDiffusionXLControlNetXSPipeline",
]
)
_import_structure["deepfloyd_if"] = [
"IFImg2ImgPipeline",
"IFImg2ImgSuperResolutionPipeline",
@@ -378,6 +384,10 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
StableDiffusionXLControlNetInpaintPipeline,
StableDiffusionXLControlNetPipeline,
)
from .controlnet_xs import (
StableDiffusionControlNetXSPipeline,
StableDiffusionXLControlNetXSPipeline,
)
from .deepfloyd_if import (
IFImg2ImgPipeline,
IFImg2ImgSuperResolutionPipeline,

68
src/diffusers/pipelines/controlnet_xs/__init__.py Normal file
View File

@@ -0,0 +1,68 @@
from typing import TYPE_CHECKING
from ...utils import (
DIFFUSERS_SLOW_IMPORT,
OptionalDependencyNotAvailable,
_LazyModule,
get_objects_from_module,
is_flax_available,
is_torch_available,
is_transformers_available,
)
_dummy_objects = {}
_import_structure = {}
try:
if not (is_transformers_available() and is_torch_available()):
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
from ...utils import dummy_torch_and_transformers_objects # noqa F403
_dummy_objects.update(get_objects_from_module(dummy_torch_and_transformers_objects))
else:
_import_structure["pipeline_controlnet_xs"] = ["StableDiffusionControlNetXSPipeline"]
_import_structure["pipeline_controlnet_xs_sd_xl"] = ["StableDiffusionXLControlNetXSPipeline"]
try:
if not (is_transformers_available() and is_flax_available()):
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
from ...utils import dummy_flax_and_transformers_objects # noqa F403
_dummy_objects.update(get_objects_from_module(dummy_flax_and_transformers_objects))
else:
pass # _import_structure["pipeline_flax_controlnet"] = ["FlaxStableDiffusionControlNetPipeline"]
if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
try:
if not (is_transformers_available() and is_torch_available()):
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
from ...utils.dummy_torch_and_transformers_objects import *
else:
from .pipeline_controlnet_xs import StableDiffusionControlNetXSPipeline
from .pipeline_controlnet_xs_sd_xl import StableDiffusionXLControlNetXSPipeline
try:
if not (is_transformers_available() and is_flax_available()):
raise OptionalDependencyNotAvailable()
except OptionalDependencyNotAvailable:
from ...utils.dummy_flax_and_transformers_objects import * # noqa F403
else:
pass # from .pipeline_flax_controlnet import FlaxStableDiffusionControlNetPipeline
else:
import sys
sys.modules[__name__] = _LazyModule(
__name__,
globals()["__file__"],
_import_structure,
module_spec=__spec__,
)
for name, value in _dummy_objects.items():
setattr(sys.modules[__name__], name, value)

900
src/diffusers/pipelines/controlnet_xs/pipeline_controlnet_xs.py Normal file
View File

@@ -0,0 +1,900 @@
# 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 Any, Callable, Dict, List, Optional, Union
import numpy as np
import PIL.Image
import torch
import torch.nn.functional as F
from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer
from ...image_processor import PipelineImageInput, VaeImageProcessor
from ...loaders import FromSingleFileMixin, LoraLoaderMixin, TextualInversionLoaderMixin
from ...models import AutoencoderKL, ControlNetXSAdapter, UNet2DConditionModel, UNetControlNetXSModel
from ...models.lora import adjust_lora_scale_text_encoder
from ...schedulers import KarrasDiffusionSchedulers
from ...utils import (
USE_PEFT_BACKEND,
deprecate,
logging,
replace_example_docstring,
scale_lora_layers,
unscale_lora_layers,
)
from ...utils.torch_utils import is_compiled_module, is_torch_version, randn_tensor
from ..pipeline_utils import DiffusionPipeline, StableDiffusionMixin
from ..stable_diffusion.pipeline_output import StableDiffusionPipelineOutput
from ..stable_diffusion.safety_checker import StableDiffusionSafetyChecker
logger = logging.get_logger(__name__) # pylint: disable=invalid-name
EXAMPLE_DOC_STRING = """
Examples:
```py
>>> # !pip install opencv-python transformers accelerate
>>> from diffusers import StableDiffusionControlNetXSPipeline, ControlNetXSAdapter
>>> from diffusers.utils import load_image
>>> import numpy as np
>>> import torch
>>> import cv2
>>> from PIL import Image
>>> prompt = "aerial view, a futuristic research complex in a bright foggy jungle, hard lighting"
>>> negative_prompt = "low quality, bad quality, sketches"
>>> # download an image
>>> image = load_image(
... "https://hf.co/datasets/hf-internal-testing/diffusers-images/resolve/main/sd_controlnet/hf-logo.png"
... )
>>> # initialize the models and pipeline
>>> controlnet_conditioning_scale = 0.5
>>> controlnet = ControlNetXSAdapter.from_pretrained(
... "UmerHA/Testing-ConrolNetXS-SD2.1-canny", torch_dtype=torch.float16
... )
>>> pipe = StableDiffusionControlNetXSPipeline.from_pretrained(
... "stabilityai/stable-diffusion-2-1-base", controlnet=controlnet, torch_dtype=torch.float16
... )
>>> pipe.enable_model_cpu_offload()
>>> # get canny image
>>> image = np.array(image)
>>> image = cv2.Canny(image, 100, 200)
>>> image = image[:, :, None]
>>> image = np.concatenate([image, image, image], axis=2)
>>> canny_image = Image.fromarray(image)
>>> # generate image
>>> image = pipe(
... prompt, controlnet_conditioning_scale=controlnet_conditioning_scale, image=canny_image
... ).images[0]
```
"""
class StableDiffusionControlNetXSPipeline(
DiffusionPipeline, StableDiffusionMixin, TextualInversionLoaderMixin, LoraLoaderMixin, FromSingleFileMixin
):
r"""
Pipeline for text-to-image generation using Stable Diffusion with ControlNet-XS guidance.
This model inherits from [`DiffusionPipeline`]. Check the superclass documentation for the generic methods
implemented for all pipelines (downloading, 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.LoraLoaderMixin.load_lora_weights`] for loading LoRA weights
- [`~loaders.LoraLoaderMixin.save_lora_weights`] for saving LoRA weights
- [`loaders.FromSingleFileMixin.from_single_file`] for loading `.ckpt` files
Args:
vae ([`AutoencoderKL`]):
Variational Auto-Encoder (VAE) model to encode and decode images to and from latent representations.
text_encoder ([`~transformers.CLIPTextModel`]):
Frozen text-encoder ([clip-vit-large-patch14](https://huggingface.co/openai/clip-vit-large-patch14)).
tokenizer ([`~transformers.CLIPTokenizer`]):
A `CLIPTokenizer` to tokenize text.
unet ([`UNet2DConditionModel`]):
A [`UNet2DConditionModel`] used to create a UNetControlNetXSModel to denoise the encoded image latents.
controlnet ([`ControlNetXSAdapter`]):
A [`ControlNetXSAdapter`] to be used in combination with `unet` 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 more details
about a model's potential harms.
feature_extractor ([`~transformers.CLIPImageProcessor`]):
A `CLIPImageProcessor` to extract features from generated images; used as inputs to the `safety_checker`.
"""
model_cpu_offload_seq = "text_encoder->unet->vae"
_optional_components = ["safety_checker", "feature_extractor"]
_exclude_from_cpu_offload = ["safety_checker"]
_callback_tensor_inputs = ["latents", "prompt_embeds", "negative_prompt_embeds"]
def __init__(
self,
vae: AutoencoderKL,
text_encoder: CLIPTextModel,
tokenizer: CLIPTokenizer,
unet: Union[UNet2DConditionModel, UNetControlNetXSModel],
controlnet: ControlNetXSAdapter,
scheduler: KarrasDiffusionSchedulers,
safety_checker: StableDiffusionSafetyChecker,
feature_extractor: CLIPImageProcessor,
requires_safety_checker: bool = True,
):
super().__init__()
if isinstance(unet, UNet2DConditionModel):
unet = UNetControlNetXSModel.from_unet(unet, controlnet)
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."
)
self.register_modules(
vae=vae,
text_encoder=text_encoder,
tokenizer=tokenizer,
unet=unet,
controlnet=controlnet,
scheduler=scheduler,
safety_checker=safety_checker,
feature_extractor=feature_extractor,
)
self.vae_scale_factor = 2 ** (len(self.vae.config.block_out_channels) - 1)
self.image_processor = VaeImageProcessor(vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True)
self.control_image_processor = VaeImageProcessor(
vae_scale_factor=self.vae_scale_factor, do_convert_rgb=True, do_normalize=False
)
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.FloatTensor] = None,
negative_prompt_embeds: Optional[torch.FloatTensor] = None,
lora_scale: Optional[float] = None,
**kwargs,
):
deprecation_message = "`_encode_prompt()` is deprecated and it will be removed in a future version. Use `encode_prompt()` instead. Also, be aware that the output format changed from a concatenated tensor to a tuple."
deprecate("_encode_prompt()", "1.0.0", deprecation_message, standard_warn=False)
prompt_embeds_tuple = self.encode_prompt(
prompt=prompt,
device=device,
num_images_per_prompt=num_images_per_prompt,
do_classifier_free_guidance=do_classifier_free_guidance,
negative_prompt=negative_prompt,
prompt_embeds=prompt_embeds,
negative_prompt_embeds=negative_prompt_embeds,
lora_scale=lora_scale,
**kwargs,
)
# concatenate for backwards comp
prompt_embeds = torch.cat([prompt_embeds_tuple[1], prompt_embeds_tuple[0]])
return prompt_embeds
# 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.FloatTensor] = None,
negative_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
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.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.
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.
"""
# 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, LoraLoaderMixin):
self._lora_scale = lora_scale
# dynamically adjust the LoRA scale
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 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
if clip_skip is None:
prompt_embeds = self.text_encoder(text_input_ids.to(device), attention_mask=attention_mask)
prompt_embeds = prompt_embeds[0]
else:
prompt_embeds = self.text_encoder(
text_input_ids.to(device), attention_mask=attention_mask, output_hidden_states=True
)
# Access the `hidden_states` first, that contains a tuple of
# all the hidden states from the encoder layers. Then index into
# the tuple to access the hidden states from the desired layer.
prompt_embeds = prompt_embeds[-1][-(clip_skip + 1)]
# We also need to apply the final LayerNorm here to not mess with the
# representations. The `last_hidden_states` that we typically use for
# obtaining the final prompt representations passes through the LayerNorm
# layer.
prompt_embeds = self.text_encoder.text_model.final_layer_norm(prompt_embeds)
if self.text_encoder is not None:
prompt_embeds_dtype = self.text_encoder.dtype
elif self.unet is not None:
prompt_embeds_dtype = self.unet.dtype
else:
prompt_embeds_dtype = prompt_embeds.dtype
prompt_embeds = prompt_embeds.to(dtype=prompt_embeds_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 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 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=prompt_embeds_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)
if isinstance(self, LoraLoaderMixin) and USE_PEFT_BACKEND:
# Retrieve the original scale by scaling back the LoRA layers
unscale_lora_layers(self.text_encoder, lora_scale)
return prompt_embeds, negative_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 None:
has_nsfw_concept = None
else:
if torch.is_tensor(image):
feature_extractor_input = self.image_processor.postprocess(image, output_type="pil")
else:
feature_extractor_input = self.image_processor.numpy_to_pil(image)
safety_checker_input = self.feature_extractor(feature_extractor_input, return_tensors="pt").to(device)
image, has_nsfw_concept = self.safety_checker(
images=image, clip_input=safety_checker_input.pixel_values.to(dtype)
)
return image, has_nsfw_concept
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.decode_latents
def decode_latents(self, latents):
deprecation_message = "The decode_latents method is deprecated and will be removed in 1.0.0. Please use VaeImageProcessor.postprocess(...) instead"
deprecate("decode_latents", "1.0.0", deprecation_message, standard_warn=False)
latents = 1 / self.vae.config.scaling_factor * latents
image = self.vae.decode(latents, return_dict=False)[0]
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.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_inputs(
self,
prompt,
image,
negative_prompt=None,
prompt_embeds=None,
negative_prompt_embeds=None,
controlnet_conditioning_scale=1.0,
control_guidance_start=0.0,
control_guidance_end=1.0,
callback_on_step_end_tensor_inputs=None,
):
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 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}."
)
# Check `image` and `controlnet_conditioning_scale`
is_compiled = hasattr(F, "scaled_dot_product_attention") and isinstance(
self.unet, torch._dynamo.eval_frame.OptimizedModule
)
if (
isinstance(self.unet, UNetControlNetXSModel)
or is_compiled
and isinstance(self.unet._orig_mod, UNetControlNetXSModel)
):
self.check_image(image, prompt, prompt_embeds)
if not isinstance(controlnet_conditioning_scale, float):
raise TypeError("For single controlnet: `controlnet_conditioning_scale` must be type `float`.")
else:
assert False
start, end = control_guidance_start, control_guidance_end
if start >= end:
raise ValueError(
f"control guidance start: {start} cannot be larger or equal to control guidance end: {end}."
)
if start < 0.0:
raise ValueError(f"control guidance start: {start} can't be smaller than 0.")
if end > 1.0:
raise ValueError(f"control guidance end: {end} can't be larger than 1.0.")
def check_image(self, image, prompt, prompt_embeds):
image_is_pil = isinstance(image, PIL.Image.Image)
image_is_tensor = isinstance(image, torch.Tensor)
image_is_np = isinstance(image, np.ndarray)
image_is_pil_list = isinstance(image, list) and isinstance(image[0], PIL.Image.Image)
image_is_tensor_list = isinstance(image, list) and isinstance(image[0], torch.Tensor)
image_is_np_list = isinstance(image, list) and isinstance(image[0], np.ndarray)
if (
not image_is_pil
and not image_is_tensor
and not image_is_np
and not image_is_pil_list
and not image_is_tensor_list
and not image_is_np_list
):
raise TypeError(
f"image must be passed and be one of PIL image, numpy array, torch tensor, list of PIL images, list of numpy arrays or list of torch tensors, but is {type(image)}"
)
if image_is_pil:
image_batch_size = 1
else:
image_batch_size = len(image)
if prompt is not None and isinstance(prompt, str):
prompt_batch_size = 1
elif prompt is not None and isinstance(prompt, list):
prompt_batch_size = len(prompt)
elif prompt_embeds is not None:
prompt_batch_size = prompt_embeds.shape[0]
if image_batch_size != 1 and image_batch_size != prompt_batch_size:
raise ValueError(
f"If image batch size is not 1, image batch size must be same as prompt batch size. image batch size: {image_batch_size}, prompt batch size: {prompt_batch_size}"
)
def prepare_image(
self,
image,
width,
height,
batch_size,
num_images_per_prompt,
device,
dtype,
do_classifier_free_guidance=False,
):
image = self.control_image_processor.preprocess(image, height=height, width=width).to(dtype=torch.float32)
image_batch_size = image.shape[0]
if image_batch_size == 1:
repeat_by = batch_size
else:
# image batch size is the same as prompt batch size
repeat_by = num_images_per_prompt
image = image.repeat_interleave(repeat_by, dim=0)
image = image.to(device=device, dtype=dtype)
if do_classifier_free_guidance:
image = torch.cat([image] * 2)
return image
# 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
@property
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.guidance_scale
def guidance_scale(self):
return self._guidance_scale
@property
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.clip_skip
def clip_skip(self):
return self._clip_skip
@property
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.do_classifier_free_guidance
def do_classifier_free_guidance(self):
return self._guidance_scale > 1 and self.unet.config.time_cond_proj_dim is None
@property
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.cross_attention_kwargs
def cross_attention_kwargs(self):
return self._cross_attention_kwargs
@property
# Copied from diffusers.pipelines.stable_diffusion.pipeline_stable_diffusion.StableDiffusionPipeline.num_timesteps
def num_timesteps(self):
return self._num_timesteps
@torch.no_grad()
@replace_example_docstring(EXAMPLE_DOC_STRING)
def __call__(
self,
prompt: Union[str, List[str]] = None,
image: PipelineImageInput = None,
height: Optional[int] = None,
width: Optional[int] = None,
num_inference_steps: int = 50,
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.FloatTensor] = None,
prompt_embeds: Optional[torch.FloatTensor] = None,
negative_prompt_embeds: Optional[torch.FloatTensor] = None,
output_type: Optional[str] = "pil",
return_dict: bool = True,
cross_attention_kwargs: Optional[Dict[str, Any]] = None,
controlnet_conditioning_scale: Union[float, List[float]] = 1.0,
control_guidance_start: float = 0.0,
control_guidance_end: float = 1.0,
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"],
):
r"""
The call function to the pipeline for generation.
Args:
prompt (`str` or `List[str]`, *optional*):
The prompt or prompts to guide image generation. If not defined, you need to pass `prompt_embeds`.
image (`torch.FloatTensor`, `PIL.Image.Image`, `np.ndarray`, `List[torch.FloatTensor]`, `List[PIL.Image.Image]`, `List[np.ndarray]`,:
`List[List[torch.FloatTensor]]`, `List[List[np.ndarray]]` or `List[List[PIL.Image.Image]]`):
The ControlNet input condition to provide guidance to the `unet` for generation. If the type is
specified as `torch.FloatTensor`, it is passed to ControlNet as is. `PIL.Image.Image` can also be
accepted as an image. The dimensions of the output image defaults to `image`'s dimensions. If height
and/or width are passed, `image` is resized accordingly. If multiple ControlNets are specified in
`init`, images must be passed as a list such that each element of the list can be correctly batched for
input to a single ControlNet.
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.
guidance_scale (`float`, *optional*, defaults to 7.5):
A higher guidance scale value encourages the model to generate images closely linked to the text
`prompt` at the expense of lower image quality. Guidance scale is enabled when `guidance_scale > 1`.
negative_prompt (`str` or `List[str]`, *optional*):
The prompt or prompts to guide what to not include in image generation. If not defined, you need to
pass `negative_prompt_embeds` instead. Ignored when not using guidance (`guidance_scale < 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) from the [DDIM](https://arxiv.org/abs/2010.02502) paper. Only applies
to the [`~schedulers.DDIMScheduler`], and is ignored in other schedulers.
generator (`torch.Generator` or `List[torch.Generator]`, *optional*):
A [`torch.Generator`](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 is 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 (prompt weighting). If not
provided, text embeddings are generated from the `prompt` input argument.
negative_prompt_embeds (`torch.FloatTensor`, *optional*):
Pre-generated negative text embeddings. Can be used to easily tweak text inputs (prompt weighting). If
not provided, `negative_prompt_embeds` are generated from the `negative_prompt` input argument.
output_type (`str`, *optional*, defaults to `"pil"`):
The output format of the generated image. Choose between `PIL.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.
cross_attention_kwargs (`dict`, *optional*):
A kwargs dictionary that if specified is passed along to the [`AttentionProcessor`] as defined in
[`self.processor`](https://github.com/huggingface/diffusers/blob/main/src/diffusers/models/attention_processor.py).
controlnet_conditioning_scale (`float` or `List[float]`, *optional*, defaults to 1.0):
The outputs of the ControlNet are multiplied by `controlnet_conditioning_scale` before they are added
to the residual in the original `unet`. If multiple ControlNets are specified in `init`, you can set
the corresponding scale as a list.
control_guidance_start (`float` or `List[float]`, *optional*, defaults to 0.0):
The percentage of total steps at which the ControlNet starts applying.
control_guidance_end (`float` or `List[float]`, *optional*, defaults to 1.0):
The percentage of total steps at which the ControlNet stops applying.
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 pipeine class.
Examples:
Returns:
[`~pipelines.stable_diffusion.StableDiffusionPipelineOutput`] or `tuple`:
If `return_dict` is `True`, [`~pipelines.stable_diffusion.StableDiffusionPipelineOutput`] is returned,
otherwise a `tuple` is returned where the first element is a list with the generated images and the
second element is a list of `bool`s indicating whether the corresponding generated image contains
"not-safe-for-work" (nsfw) content.
"""
unet = self.unet._orig_mod if is_compiled_module(self.unet) else self.unet
# 1. Check inputs. Raise error if not correct
self.check_inputs(
prompt,
image,
negative_prompt,
prompt_embeds,
negative_prompt_embeds,
controlnet_conditioning_scale,
control_guidance_start,
control_guidance_end,
callback_on_step_end_tensor_inputs,
)
self._guidance_scale = guidance_scale
self._clip_skip = clip_skip
self._cross_attention_kwargs = cross_attention_kwargs
self._interrupt = False
# 2. Define call parameters
if prompt is not None and isinstance(prompt, str):
batch_size = 1
elif prompt is not None and isinstance(prompt, list):
batch_size = len(prompt)
else:
batch_size = prompt_embeds.shape[0]
device = self._execution_device
# 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
text_encoder_lora_scale = (
cross_attention_kwargs.get("scale", None) if cross_attention_kwargs is not None else None
)
prompt_embeds, negative_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,
lora_scale=text_encoder_lora_scale,
clip_skip=clip_skip,
)
# 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
if do_classifier_free_guidance:
prompt_embeds = torch.cat([negative_prompt_embeds, prompt_embeds])
# 4. Prepare image
image = self.prepare_image(
image=image,
width=width,
height=height,
batch_size=batch_size * num_images_per_prompt,
num_images_per_prompt=num_images_per_prompt,
device=device,
dtype=unet.dtype,
do_classifier_free_guidance=do_classifier_free_guidance,
)
height, width = image.shape[-2:]
# 5. Prepare timesteps
self.scheduler.set_timesteps(num_inference_steps, device=device)
timesteps = self.scheduler.timesteps
# 6. Prepare latent variables
num_channels_latents = self.unet.in_channels
latents = self.prepare_latents(
batch_size * num_images_per_prompt,
num_channels_latents,
height,
width,
prompt_embeds.dtype,
device,
generator,
latents,
)
# 7. 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)
# 8. Denoising loop
num_warmup_steps = len(timesteps) - num_inference_steps * self.scheduler.order
self._num_timesteps = len(timesteps)
is_controlnet_compiled = is_compiled_module(self.unet)
is_torch_higher_equal_2_1 = is_torch_version(">=", "2.1")
with self.progress_bar(total=num_inference_steps) as progress_bar:
for i, t in enumerate(timesteps):
# Relevant thread:
# https://dev-discuss.pytorch.org/t/cudagraphs-in-pytorch-2-0/1428
if is_controlnet_compiled and is_torch_higher_equal_2_1:
torch._inductor.cudagraph_mark_step_begin()
# expand the latents if we are doing classifier free guidance
latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents
latent_model_input = self.scheduler.scale_model_input(latent_model_input, t)
# predict the noise residual
apply_control = (
i / len(timesteps) >= control_guidance_start and (i + 1) / len(timesteps) <= control_guidance_end
)
noise_pred = self.unet(
sample=latent_model_input,
timestep=t,
encoder_hidden_states=prompt_embeds,
controlnet_cond=image,
conditioning_scale=controlnet_conditioning_scale,
cross_attention_kwargs=cross_attention_kwargs,
return_dict=True,
apply_control=apply_control,
).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)
latents = self.scheduler.step(noise_pred, t, latents, **extra_step_kwargs, return_dict=False)[0]
if callback_on_step_end is not None:
callback_kwargs = {}
for k in callback_on_step_end_tensor_inputs:
callback_kwargs[k] = locals()[k]
callback_outputs = callback_on_step_end(self, i, t, callback_kwargs)
latents = callback_outputs.pop("latents", latents)
prompt_embeds = callback_outputs.pop("prompt_embeds", prompt_embeds)
negative_prompt_embeds = callback_outputs.pop("negative_prompt_embeds", negative_prompt_embeds)
if i == len(timesteps) - 1 or ((i + 1) > num_warmup_steps and (i + 1) % self.scheduler.order == 0):
progress_bar.update()
# If we do sequential model offloading, let's offload unet and controlnet
# manually for max memory savings
if hasattr(self, "final_offload_hook") and self.final_offload_hook is not None:
self.unet.to("cpu")
self.controlnet.to("cpu")
torch.cuda.empty_cache()
if not output_type == "latent":
image = self.vae.decode(latents / self.vae.config.scaling_factor, return_dict=False, generator=generator)[
0
]
image, has_nsfw_concept = self.run_safety_checker(image, device, prompt_embeds.dtype)
else:
image = latents
has_nsfw_concept = None
if has_nsfw_concept is None:
do_denormalize = [True] * image.shape[0]
else:
do_denormalize = [not has_nsfw for has_nsfw in has_nsfw_concept]
image = self.image_processor.postprocess(image, output_type=output_type, do_denormalize=do_denormalize)
# Offload all models
self.maybe_free_model_hooks()
if not return_dict:
return (image, has_nsfw_concept)
return StableDiffusionPipelineOutput(images=image, nsfw_content_detected=has_nsfw_concept)

1125
src/diffusers/pipelines/controlnet_xs/pipeline_controlnet_xs_sd_xl.py Normal file
View File

File diff suppressed because it is too large Load Diff

27
src/diffusers/pipelines/deprecated/versatile_diffusion/modeling_text_unet.py
View File

@@ -2238,6 +2238,7 @@ class UNetMidBlockFlatCrossAttn(nn.Module):
self,
in_channels: int,
temb_channels: int,
out_channels: Optional[int] = None,
dropout: float = 0.0,
num_layers: int = 1,
transformer_layers_per_block: Union[int, Tuple[int]] = 1,
@@ -2245,6 +2246,7 @@ class UNetMidBlockFlatCrossAttn(nn.Module):
resnet_time_scale_shift: str = "default",
resnet_act_fn: str = "swish",
resnet_groups: int = 32,
resnet_groups_out: Optional[int] = None,
resnet_pre_norm: bool = True,
num_attention_heads: int = 1,
output_scale_factor: float = 1.0,
@@ -2256,6 +2258,10 @@ class UNetMidBlockFlatCrossAttn(nn.Module):
):
super().__init__()
out_channels = out_channels or in_channels
self.in_channels = in_channels
self.out_channels = out_channels
self.has_cross_attention = True
self.num_attention_heads = num_attention_heads
resnet_groups = resnet_groups if resnet_groups is not None else min(in_channels // 4, 32)
@@ -2264,14 +2270,17 @@ class UNetMidBlockFlatCrossAttn(nn.Module):
if isinstance(transformer_layers_per_block, int):
transformer_layers_per_block = [transformer_layers_per_block] * num_layers
resnet_groups_out = resnet_groups_out or resnet_groups
# there is always at least one resnet
resnets = [
ResnetBlockFlat(
in_channels=in_channels,
out_channels=in_channels,
out_channels=out_channels,
temb_channels=temb_channels,
eps=resnet_eps,
groups=resnet_groups,
groups_out=resnet_groups_out,
dropout=dropout,
time_embedding_norm=resnet_time_scale_shift,
non_linearity=resnet_act_fn,
@@ -2286,11 +2295,11 @@ class UNetMidBlockFlatCrossAttn(nn.Module):
attentions.append(
Transformer2DModel(
num_attention_heads,
in_channels // num_attention_heads,
in_channels=in_channels,
out_channels // num_attention_heads,
in_channels=out_channels,
num_layers=transformer_layers_per_block[i],
cross_attention_dim=cross_attention_dim,
norm_num_groups=resnet_groups,
norm_num_groups=resnet_groups_out,
use_linear_projection=use_linear_projection,
upcast_attention=upcast_attention,
attention_type=attention_type,
@@ -2300,8 +2309,8 @@ class UNetMidBlockFlatCrossAttn(nn.Module):
attentions.append(
DualTransformer2DModel(
num_attention_heads,
in_channels // num_attention_heads,
in_channels=in_channels,
out_channels // num_attention_heads,
in_channels=out_channels,
num_layers=1,
cross_attention_dim=cross_attention_dim,
norm_num_groups=resnet_groups,
@@ -2309,11 +2318,11 @@ class UNetMidBlockFlatCrossAttn(nn.Module):
)
resnets.append(
ResnetBlockFlat(
in_channels=in_channels,
out_channels=in_channels,
in_channels=out_channels,
out_channels=out_channels,
temb_channels=temb_channels,
eps=resnet_eps,
groups=resnet_groups,
groups=resnet_groups_out,
dropout=dropout,
time_embedding_norm=resnet_time_scale_shift,
non_linearity=resnet_act_fn,

30
src/diffusers/utils/dummy_pt_objects.py
View File

@@ -92,6 +92,21 @@ class ControlNetModel(metaclass=DummyObject):
requires_backends(cls, ["torch"])
class ControlNetXSAdapter(metaclass=DummyObject):
_backends = ["torch"]
def __init__(self, *args, **kwargs):
requires_backends(self, ["torch"])
@classmethod
def from_config(cls, *args, **kwargs):
requires_backends(cls, ["torch"])
@classmethod
def from_pretrained(cls, *args, **kwargs):
requires_backends(cls, ["torch"])
class I2VGenXLUNet(metaclass=DummyObject):
_backends = ["torch"]
@@ -287,6 +302,21 @@ class UNet3DConditionModel(metaclass=DummyObject):
requires_backends(cls, ["torch"])
class UNetControlNetXSModel(metaclass=DummyObject):
_backends = ["torch"]
def __init__(self, *args, **kwargs):
requires_backends(self, ["torch"])
@classmethod
def from_config(cls, *args, **kwargs):
requires_backends(cls, ["torch"])
@classmethod
def from_pretrained(cls, *args, **kwargs):
requires_backends(cls, ["torch"])
class UNetMotionModel(metaclass=DummyObject):
_backends = ["torch"]

30
src/diffusers/utils/dummy_torch_and_transformers_objects.py
View File

@@ -902,6 +902,21 @@ class StableDiffusionControlNetPipeline(metaclass=DummyObject):
requires_backends(cls, ["torch", "transformers"])
class StableDiffusionControlNetXSPipeline(metaclass=DummyObject):
_backends = ["torch", "transformers"]
def __init__(self, *args, **kwargs):
requires_backends(self, ["torch", "transformers"])
@classmethod
def from_config(cls, *args, **kwargs):
requires_backends(cls, ["torch", "transformers"])
@classmethod
def from_pretrained(cls, *args, **kwargs):
requires_backends(cls, ["torch", "transformers"])
class StableDiffusionDepth2ImgPipeline(metaclass=DummyObject):
_backends = ["torch", "transformers"]
@@ -1247,6 +1262,21 @@ class StableDiffusionXLControlNetPipeline(metaclass=DummyObject):
requires_backends(cls, ["torch", "transformers"])
class StableDiffusionXLControlNetXSPipeline(metaclass=DummyObject):
_backends = ["torch", "transformers"]
def __init__(self, *args, **kwargs):
requires_backends(self, ["torch", "transformers"])
@classmethod
def from_config(cls, *args, **kwargs):
requires_backends(cls, ["torch", "transformers"])
@classmethod
def from_pretrained(cls, *args, **kwargs):
requires_backends(cls, ["torch", "transformers"])
class StableDiffusionXLImg2ImgPipeline(metaclass=DummyObject):
_backends = ["torch", "transformers"]