Move ControlNetXS into Community Folder

docs/source/en/_toctree.yml

@@ -266,10 +266,6 @@
       title: ControlNet
     - local: api/pipelines/controlnet_sdxl
       title: ControlNet with Stable Diffusion XL
-    - local: api/pipelines/controlnetxs
-      title: ControlNet-XS
-    - local: api/pipelines/controlnetxs_sdxl
-      title: ControlNet-XS with Stable Diffusion XL
     - local: api/pipelines/dance_diffusion
       title: Dance Diffusion
     - local: api/pipelines/ddim

docs/source/en/api/pipelines/controlnetxs.md → examples/research_projects/controlnetxs/README.md

@@ -1,15 +1,3 @@
-<!--Copyright 2023 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.
--->
-
 # ControlNet-XS
 
 ControlNet-XS was introduced in [ControlNet-XS](https://vislearn.github.io/ControlNet-XS/) by Denis Zavadski and Carsten Rother. It is based on the observation that the control model in the [original ControlNet](https://huggingface.co/papers/2302.05543) can be made much smaller and still produce good results.
@@ -24,16 +12,5 @@ Here's the overview from the [project page](https://vislearn.github.io/ControlNe
 
 This model was contributed by [UmerHA](https://twitter.com/UmerHAdil). ❤️
 
-<Tip>
-
-Make sure to check out the Schedulers [guide](../../using-diffusers/schedulers) to learn how to explore the tradeoff between scheduler speed and quality, and see the [reuse components across pipelines](../../using-diffusers/loading#reuse-components-across-pipelines) section to learn how to efficiently load the same components into multiple pipelines.
-
-</Tip>
-
-## StableDiffusionControlNetXSPipeline
-[[autodoc]] StableDiffusionControlNetXSPipeline
-	- all
-	- __call__
 
-## StableDiffusionPipelineOutput
-[[autodoc]] pipelines.stable_diffusion.StableDiffusionPipelineOutput
+> 🧠 Make sure to check out the Schedulers [guide](https://huggingface.co/docs/diffusers/main/en/using-diffusers/schedulers) to learn how to explore the tradeoff between scheduler speed and quality, and see the [reuse components across pipelines](https://huggingface.co/docs/diffusers/main/en/using-diffusers/loading#reuse-components-across-pipelines) section to learn how to efficiently load the same components into multiple pipelines.

docs/source/en/api/pipelines/controlnetxs_sdxl.md → examples/research_projects/controlnetxs/README_sdxl.md

@@ -1,15 +1,3 @@
-<!--Copyright 2023 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.
--->
-
 # ControlNet-XS with Stable Diffusion XL
 
 ControlNet-XS was introduced in [ControlNet-XS](https://vislearn.github.io/ControlNet-XS/) by Denis Zavadski and Carsten Rother. It is based on the observation that the control model in the [original ControlNet](https://huggingface.co/papers/2302.05543) can be made much smaller and still produce good results.
@@ -24,22 +12,4 @@ Here's the overview from the [project page](https://vislearn.github.io/ControlNe
 
 This model was contributed by [UmerHA](https://twitter.com/UmerHAdil). ❤️
 
-<Tip warning={true}>
-
-🧪 Many of the SDXL ControlNet checkpoints are experimental, and there is a lot of room for improvement. Feel free to open an [Issue](https://github.com/huggingface/diffusers/issues/new/choose) and leave us feedback on how we can improve!
-
-</Tip>
-
-<Tip>
-
-Make sure to check out the Schedulers [guide](../../using-diffusers/schedulers) to learn how to explore the tradeoff between scheduler speed and quality, and see the [reuse components across pipelines](../../using-diffusers/loading#reuse-components-across-pipelines) section to learn how to efficiently load the same components into multiple pipelines.
-
-</Tip>
-
-## StableDiffusionXLControlNetXSPipeline
-[[autodoc]] StableDiffusionXLControlNetXSPipeline
-	- all
-	- __call__
-
-## StableDiffusionPipelineOutput
-[[autodoc]] pipelines.stable_diffusion.StableDiffusionPipelineOutput
+> 🧠 Make sure to check out the Schedulers [guide](https://huggingface.co/docs/diffusers/main/en/using-diffusers/schedulers) to learn how to explore the tradeoff between scheduler speed and quality, and see the [reuse components across pipelines](https://huggingface.co/docs/diffusers/main/en/using-diffusers/loading#reuse-components-across-pipelines) section to learn how to efficiently load the same components into multiple pipelines.

src/diffusers/models/controlnetxs.py → examples/research_projects/controlnetxs/controlnetxs.py

@@ -21,13 +21,12 @@
 from torch.nn import functional as F
 from torch.nn.modules.normalization import GroupNorm
 
-from ..configuration_utils import ConfigMixin, register_to_config
-from ..utils import BaseOutput, logging
-from .attention_processor import USE_PEFT_BACKEND, AttentionProcessor
-from .autoencoders import AutoencoderKL
-from .lora import LoRACompatibleConv
-from .modeling_utils import ModelMixin
-from .unet_2d_blocks import (
+from diffusers.configuration_utils import ConfigMixin, register_to_config
+from diffusers.models.attention_processor import USE_PEFT_BACKEND, AttentionProcessor
+from diffusers.models.autoencoders import AutoencoderKL
+from diffusers.models.lora import LoRACompatibleConv
+from diffusers.models.modeling_utils import ModelMixin
+from diffusers.models.unet_2d_blocks import (
     CrossAttnDownBlock2D,
     CrossAttnUpBlock2D,
     DownBlock2D,
@@ -37,7 +36,8 @@
     UpBlock2D,
     Upsample2D,
 )
-from .unet_2d_condition import UNet2DConditionModel
+from diffusers.models.unet_2d_condition import UNet2DConditionModel
+from diffusers.utils import BaseOutput, logging
 
 
 logger = logging.get_logger(__name__)  # pylint: disable=invalid-name

examples/research_projects/controlnetxs/infer_sd_controlnetxs.py

@@ -0,0 +1,58 @@
+# !pip install opencv-python transformers accelerate
+import argparse
+
+import cv2
+import numpy as np
+import torch
+from controlnetxs import ControlNetXSModel
+from PIL import Image
+from pipeline_controlnet_xs import StableDiffusionControlNetXSPipeline
+
+from diffusers.utils import load_image
+
+
+parser = argparse.ArgumentParser()
+parser.add_argument(
+    "--prompt", type=str, default="aerial view, a futuristic research complex in a bright foggy jungle, hard lighting"
+)
+parser.add_argument("--negative_prompt", type=str, default="low quality, bad quality, sketches")
+parser.add_argument("--controlnet_conditioning_scale", type=float, default=0.7)
+parser.add_argument(
+    "--image_path",
+    type=str,
+    default="https://hf.co/datasets/hf-internal-testing/diffusers-images/resolve/main/sd_controlnet/hf-logo.png",
+)
+parser.add_argument("--num_inference_steps", type=int, default=50)
+
+args = parser.parse_args()
+
+prompt = args.prompt
+negative_prompt = args.negative_prompt
+# download an image
+image = load_image(args.image_path)
+
+# initialize the models and pipeline
+controlnet_conditioning_scale = args.controlnet_conditioning_scale
+controlnet = ControlNetXSModel.from_pretrained("UmerHA/ConrolNetXS-SD2.1-canny", torch_dtype=torch.float16)
+pipe = StableDiffusionControlNetXSPipeline.from_pretrained(
+    "stabilityai/stable-diffusion-2-1", 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)
+
+num_inference_steps = args.num_inference_steps
+
+# generate image
+image = pipe(
+    prompt,
+    controlnet_conditioning_scale=controlnet_conditioning_scale,
+    image=canny_image,
+    num_inference_steps=num_inference_steps,
+).images[0]
+image.save("cnxs_sd.canny.png")

examples/research_projects/controlnetxs/infer_sdxl_controlnetxs.py

@@ -0,0 +1,57 @@
+# !pip install opencv-python transformers accelerate
+import argparse
+
+import cv2
+import numpy as np
+import torch
+from controlnetxs import ControlNetXSModel
+from PIL import Image
+from pipeline_controlnet_xs import StableDiffusionControlNetXSPipeline
+
+from diffusers.utils import load_image
+
+
+parser = argparse.ArgumentParser()
+parser.add_argument(
+    "--prompt", type=str, default="aerial view, a futuristic research complex in a bright foggy jungle, hard lighting"
+)
+parser.add_argument("--negative_prompt", type=str, default="low quality, bad quality, sketches")
+parser.add_argument("--controlnet_conditioning_scale", type=float, default=0.7)
+parser.add_argument(
+    "--image_path",
+    type=str,
+    default="https://hf.co/datasets/hf-internal-testing/diffusers-images/resolve/main/sd_controlnet/hf-logo.png",
+)
+parser.add_argument("--num_inference_steps", type=int, default=50)
+
+args = parser.parse_args()
+
+prompt = args.prompt
+negative_prompt = args.negative_prompt
+# download an image
+image = load_image(args.image_path)
+# initialize the models and pipeline
+controlnet_conditioning_scale = args.controlnet_conditioning_scale
+controlnet = ControlNetXSModel.from_pretrained("UmerHA/ConrolNetXS-SDXL-canny", torch_dtype=torch.float16)
+pipe = StableDiffusionControlNetXSPipeline.from_pretrained(
+    "stabilityai/stable-diffusion-xl-base-1.0", 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)
+
+num_inference_steps = args.num_inference_steps
+
+# generate image
+image = pipe(
+    prompt,
+    controlnet_conditioning_scale=controlnet_conditioning_scale,
+    image=canny_image,
+    num_inference_steps=num_inference_steps,
+).images[0]
+image.save("cnxs_sdxl.canny.png")

src/diffusers/pipelines/controlnet_xs/pipeline_controlnet_xs.py → examples/research_projects/controlnetxs/pipeline_controlnet_xs.py

@@ -19,74 +19,30 @@
 import PIL.Image
 import torch
 import torch.nn.functional as F
+from controlnetxs import ControlNetXSModel
 from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer
 
-from ...image_processor import PipelineImageInput, VaeImageProcessor
-from ...loaders import FromSingleFileMixin, LoraLoaderMixin, TextualInversionLoaderMixin
-from ...models import AutoencoderKL, ControlNetXSModel, UNet2DConditionModel
-from ...models.lora import adjust_lora_scale_text_encoder
-from ...schedulers import KarrasDiffusionSchedulers
-from ...utils import (
+from diffusers.image_processor import PipelineImageInput, VaeImageProcessor
+from diffusers.loaders import FromSingleFileMixin, LoraLoaderMixin, TextualInversionLoaderMixin
+from diffusers.models import AutoencoderKL, UNet2DConditionModel
+from diffusers.models.lora import adjust_lora_scale_text_encoder
+from diffusers.pipelines.pipeline_utils import DiffusionPipeline
+from diffusers.pipelines.stable_diffusion.pipeline_output import StableDiffusionPipelineOutput
+from diffusers.pipelines.stable_diffusion.safety_checker import StableDiffusionSafetyChecker
+from diffusers.schedulers import KarrasDiffusionSchedulers
+from diffusers.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
-from ..stable_diffusion.pipeline_output import StableDiffusionPipelineOutput
-from ..stable_diffusion.safety_checker import StableDiffusionSafetyChecker
+from diffusers.utils.torch_utils import is_compiled_module, is_torch_version, randn_tensor
 
 
 logger = logging.get_logger(__name__)  # pylint: disable=invalid-name
 
 
-EXAMPLE_DOC_STRING = """
-    Examples:
-        ```py
-        >>> # !pip install opencv-python transformers accelerate
-        >>> from diffusers import StableDiffusionControlNetXSPipeline, ControlNetXSModel
-        >>> 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 = ControlNetXSModel.from_pretrained(
-        ...     "UmerHA/ConrolNetXS-SD2.1-canny", torch_dtype=torch.float16
-        ... )
-        >>> pipe = StableDiffusionControlNetXSPipeline.from_pretrained(
-        ...     "stabilityai/stable-diffusion-2-1", 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, TextualInversionLoaderMixin, LoraLoaderMixin, FromSingleFileMixin
 ):
@@ -669,7 +625,6 @@ def disable_freeu(self):
         self.unet.disable_freeu()
 
     @torch.no_grad()
-    @replace_example_docstring(EXAMPLE_DOC_STRING)
     def __call__(
         self,
         prompt: Union[str, List[str]] = None,