Update CLIPFeatureExtractor to CLIPImageProcessor and DPTFeatureExtractor to DPTImageProcessor (#9002)

* fix: update `CLIPFeatureExtractor` to `CLIPImageProcessor` in codebase

* `make style && make quality`

* Update `DPTFeatureExtractor` to `DPTImageProcessor` in codebase

* `make style`

---------

Co-authored-by: Aryan <aryan@huggingface.co>

Author: tolgacangoz

docs/source/en/using-diffusers/custom_pipeline_overview.md

@@ -289,9 +289,9 @@ scheduler = DPMSolverMultistepScheduler.from_pretrained(pipe_id, subfolder="sche
 3. Load an image processor:
 
 ```python
-from transformers import CLIPFeatureExtractor
+from transformers import CLIPImageProcessor
 
-feature_extractor = CLIPFeatureExtractor.from_pretrained(pipe_id, subfolder="feature_extractor")
+feature_extractor = CLIPImageProcessor.from_pretrained(pipe_id, subfolder="feature_extractor")
 ```
 
 <Tip warning={true}>

docs/source/en/using-diffusers/inference_with_tcd_lora.md

@@ -212,14 +212,14 @@ TCD-LoRA is very versatile, and it can be combined with other adapter types like
 import torch
 import numpy as np
 from PIL import Image
-from transformers import DPTFeatureExtractor, DPTForDepthEstimation
+from transformers import DPTImageProcessor, DPTForDepthEstimation
 from diffusers import ControlNetModel, StableDiffusionXLControlNetPipeline
 from diffusers.utils import load_image, make_image_grid
 from scheduling_tcd import TCDScheduler
 
 device = "cuda"
 depth_estimator = DPTForDepthEstimation.from_pretrained("Intel/dpt-hybrid-midas").to(device)
-feature_extractor = DPTFeatureExtractor.from_pretrained("Intel/dpt-hybrid-midas")
+feature_extractor = DPTImageProcessor.from_pretrained("Intel/dpt-hybrid-midas")
 
 def get_depth_map(image):
     image = feature_extractor(images=image, return_tensors="pt").pixel_values.to(device)

docs/source/ko/using-diffusers/loading.md

@@ -307,7 +307,7 @@ print(pipeline)
 
 위의 코드 출력 결과를 확인해보면, `pipeline`은 [`StableDiffusionPipeline`]의 인스턴스이며, 다음과 같이 총 7개의 컴포넌트로 구성된다는 것을 알 수 있습니다.
 
-- `"feature_extractor"`: [`~transformers.CLIPFeatureExtractor`]의 인스턴스
+- `"feature_extractor"`: [`~transformers.CLIPImageProcessor`]의 인스턴스
 - `"safety_checker"`: 유해한 컨텐츠를 스크리닝하기 위한 [컴포넌트](https://github.com/huggingface/diffusers/blob/e55687e1e15407f60f32242027b7bb8170e58266/src/diffusers/pipelines/stable_diffusion/safety_checker.py#L32)
 - `"scheduler"`: [`PNDMScheduler`]의 인스턴스
 - `"text_encoder"`: [`~transformers.CLIPTextModel`]의 인스턴스

docs/source/ko/using-diffusers/textual_inversion_inference.md

@@ -24,7 +24,7 @@ import PIL
 from PIL import Image
 
 from diffusers import StableDiffusionPipeline
-from transformers import CLIPFeatureExtractor, CLIPTextModel, CLIPTokenizer
+from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer
 
 
 def image_grid(imgs, rows, cols):

examples/community/README.md

@@ -1435,9 +1435,9 @@ import requests
 import torch
 from diffusers import DiffusionPipeline
 from PIL import Image
-from transformers import CLIPFeatureExtractor, CLIPModel
+from transformers import CLIPImageProcessor, CLIPModel
 
-feature_extractor = CLIPFeatureExtractor.from_pretrained(
+feature_extractor = CLIPImageProcessor.from_pretrained(
     "laion/CLIP-ViT-B-32-laion2B-s34B-b79K"
 )
 clip_model = CLIPModel.from_pretrained(
@@ -2122,15 +2122,15 @@ import torch
 import open_clip
 from open_clip import SimpleTokenizer
 from diffusers import DiffusionPipeline
-from transformers import CLIPFeatureExtractor, CLIPModel
+from transformers import CLIPImageProcessor, CLIPModel
 
 
 def download_image(url):
     response = requests.get(url)
     return PIL.Image.open(BytesIO(response.content)).convert("RGB")
 
 # Loading additional models
-feature_extractor = CLIPFeatureExtractor.from_pretrained(
+feature_extractor = CLIPImageProcessor.from_pretrained(
     "laion/CLIP-ViT-B-32-laion2B-s34B-b79K"
 )
 clip_model = CLIPModel.from_pretrained(

examples/community/clip_guided_images_mixing_stable_diffusion.py

@@ -7,7 +7,7 @@
 import torch
 from torch.nn import functional as F
 from torchvision import transforms
-from transformers import CLIPFeatureExtractor, CLIPModel, CLIPTextModel, CLIPTokenizer
+from transformers import CLIPImageProcessor, CLIPModel, CLIPTextModel, CLIPTokenizer
 
 from diffusers import (
     AutoencoderKL,
@@ -86,7 +86,7 @@ def __init__(
         tokenizer: CLIPTokenizer,
         unet: UNet2DConditionModel,
         scheduler: Union[PNDMScheduler, LMSDiscreteScheduler, DDIMScheduler, DPMSolverMultistepScheduler],
-        feature_extractor: CLIPFeatureExtractor,
+        feature_extractor: CLIPImageProcessor,
         coca_model=None,
         coca_tokenizer=None,
         coca_transform=None,

examples/community/clip_guided_stable_diffusion_img2img.py

@@ -7,7 +7,7 @@
 from torch import nn
 from torch.nn import functional as F
 from torchvision import transforms
-from transformers import CLIPFeatureExtractor, CLIPModel, CLIPTextModel, CLIPTokenizer
+from transformers import CLIPImageProcessor, CLIPModel, CLIPTextModel, CLIPTokenizer
 
 from diffusers import (
     AutoencoderKL,
@@ -32,9 +32,9 @@
         import torch
         from diffusers import DiffusionPipeline
         from PIL import Image
-        from transformers import CLIPFeatureExtractor, CLIPModel
+        from transformers import CLIPImageProcessor, CLIPModel
 
-        feature_extractor = CLIPFeatureExtractor.from_pretrained(
+        feature_extractor = CLIPImageProcessor.from_pretrained(
             "laion/CLIP-ViT-B-32-laion2B-s34B-b79K"
         )
         clip_model = CLIPModel.from_pretrained(
@@ -139,7 +139,7 @@ def __init__(
         tokenizer: CLIPTokenizer,
         unet: UNet2DConditionModel,
         scheduler: Union[PNDMScheduler, LMSDiscreteScheduler, DDIMScheduler, DPMSolverMultistepScheduler],
-        feature_extractor: CLIPFeatureExtractor,
+        feature_extractor: CLIPImageProcessor,
     ):
         super().__init__()
         self.register_modules(

examples/community/mixture_canvas.py

@@ -9,7 +9,7 @@
 from numpy import exp, pi, sqrt
 from torchvision.transforms.functional import resize
 from tqdm.auto import tqdm
-from transformers import CLIPFeatureExtractor, CLIPTextModel, CLIPTokenizer
+from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer
 
 from diffusers.models import AutoencoderKL, UNet2DConditionModel
 from diffusers.pipelines.pipeline_utils import DiffusionPipeline, StableDiffusionMixin
@@ -275,7 +275,7 @@ def __init__(
         unet: UNet2DConditionModel,
         scheduler: Union[DDIMScheduler, LMSDiscreteScheduler, PNDMScheduler],
         safety_checker: StableDiffusionSafetyChecker,
-        feature_extractor: CLIPFeatureExtractor,
+        feature_extractor: CLIPImageProcessor,
     ):
         super().__init__()
         self.register_modules(

examples/community/mixture_tiling.py

@@ -15,7 +15,7 @@
 
 try:
     from ligo.segments import segment
-    from transformers import CLIPFeatureExtractor, CLIPTextModel, CLIPTokenizer
+    from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer
 except ImportError:
     raise ImportError("Please install transformers and ligo-segments to use the mixture pipeline")
 
@@ -144,7 +144,7 @@ def __init__(
         unet: UNet2DConditionModel,
         scheduler: Union[DDIMScheduler, PNDMScheduler],
         safety_checker: StableDiffusionSafetyChecker,
-        feature_extractor: CLIPFeatureExtractor,
+        feature_extractor: CLIPImageProcessor,
     ):
         super().__init__()
         self.register_modules(

examples/community/pipeline_stable_diffusion_xl_controlnet_adapter.py

@@ -189,7 +189,7 @@ class StableDiffusionXLControlNetAdapterPipeline(
         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 ([`CLIPFeatureExtractor`]):
+        feature_extractor ([`CLIPImageProcessor`]):
             Model that extracts features from generated images to be used as inputs for the `safety_checker`.
     """
 

examples/community/pipeline_stable_diffusion_xl_controlnet_adapter_inpaint.py

@@ -332,7 +332,7 @@ class StableDiffusionXLControlNetAdapterInpaintPipeline(
         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 ([`CLIPFeatureExtractor`]):
+        feature_extractor ([`CLIPImageProcessor`]):
             Model that extracts features from generated images to be used as inputs for the `safety_checker`.
         requires_aesthetics_score (`bool`, *optional*, defaults to `"False"`):
             Whether the `unet` requires a aesthetic_score condition to be passed during inference. Also see the config

examples/community/pipeline_zero1to3.py

@@ -9,7 +9,7 @@
 import PIL.Image
 import torch
 from packaging import version
-from transformers import CLIPFeatureExtractor, CLIPVisionModelWithProjection
+from transformers import CLIPImageProcessor, CLIPVisionModelWithProjection
 
 # from ...configuration_utils import FrozenDict
 # from ...models import AutoencoderKL, UNet2DConditionModel
@@ -87,7 +87,7 @@ class Zero1to3StableDiffusionPipeline(DiffusionPipeline, StableDiffusionMixin):
         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 ([`CLIPFeatureExtractor`]):
+        feature_extractor ([`CLIPImageProcessor`]):
             Model that extracts features from generated images to be used as inputs for the `safety_checker`.
         cc_projection ([`CCProjection`]):
             Projection layer to project the concated CLIP features and pose embeddings to the original CLIP feature size.
@@ -102,7 +102,7 @@ def __init__(
         unet: UNet2DConditionModel,
         scheduler: KarrasDiffusionSchedulers,
         safety_checker: StableDiffusionSafetyChecker,
-        feature_extractor: CLIPFeatureExtractor,
+        feature_extractor: CLIPImageProcessor,
         cc_projection: CCProjection,
         requires_safety_checker: bool = True,
     ):

examples/community/regional_prompting_stable_diffusion.py

@@ -3,7 +3,7 @@
 
 import torch
 import torchvision.transforms.functional as FF
-from transformers import CLIPFeatureExtractor, CLIPTextModel, CLIPTokenizer
+from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer
 
 from diffusers import StableDiffusionPipeline
 from diffusers.models import AutoencoderKL, UNet2DConditionModel
@@ -69,7 +69,7 @@ def __init__(
         unet: UNet2DConditionModel,
         scheduler: KarrasDiffusionSchedulers,
         safety_checker: StableDiffusionSafetyChecker,
-        feature_extractor: CLIPFeatureExtractor,
+        feature_extractor: CLIPImageProcessor,
         requires_safety_checker: bool = True,
     ):
         super().__init__(

examples/community/stable_diffusion_ipex.py

@@ -18,7 +18,7 @@
 import intel_extension_for_pytorch as ipex
 import torch
 from packaging import version
-from transformers import CLIPFeatureExtractor, CLIPTextModel, CLIPTokenizer
+from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer
 
 from diffusers.configuration_utils import FrozenDict
 from diffusers.loaders import StableDiffusionLoraLoaderMixin, TextualInversionLoaderMixin
@@ -86,7 +86,7 @@ class StableDiffusionIPEXPipeline(
         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 ([`CLIPFeatureExtractor`]):
+        feature_extractor ([`CLIPImageProcessor`]):
             Model that extracts features from generated images to be used as inputs for the `safety_checker`.
     """
 
@@ -100,7 +100,7 @@ def __init__(
         unet: UNet2DConditionModel,
         scheduler: KarrasDiffusionSchedulers,
         safety_checker: StableDiffusionSafetyChecker,
-        feature_extractor: CLIPFeatureExtractor,
+        feature_extractor: CLIPImageProcessor,
         requires_safety_checker: bool = True,
     ):
         super().__init__()

examples/community/stable_diffusion_tensorrt_img2img.py

@@ -42,7 +42,7 @@
     network_from_onnx_path,
     save_engine,
 )
-from transformers import CLIPFeatureExtractor, CLIPTextModel, CLIPTokenizer, CLIPVisionModelWithProjection
+from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer, CLIPVisionModelWithProjection
 
 from diffusers import DiffusionPipeline
 from diffusers.configuration_utils import FrozenDict, deprecate
@@ -679,7 +679,7 @@ class TensorRTStableDiffusionImg2ImgPipeline(DiffusionPipeline):
         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 ([`CLIPFeatureExtractor`]):
+        feature_extractor ([`CLIPImageProcessor`]):
             Model that extracts features from generated images to be used as inputs for the `safety_checker`.
     """
 
@@ -693,7 +693,7 @@ def __init__(
         unet: UNet2DConditionModel,
         scheduler: DDIMScheduler,
         safety_checker: StableDiffusionSafetyChecker,
-        feature_extractor: CLIPFeatureExtractor,
+        feature_extractor: CLIPImageProcessor,
         image_encoder: CLIPVisionModelWithProjection = None,
         requires_safety_checker: bool = True,
         stages=["clip", "unet", "vae", "vae_encoder"],

examples/community/stable_diffusion_tensorrt_inpaint.py

@@ -42,7 +42,7 @@
     network_from_onnx_path,
     save_engine,
 )
-from transformers import CLIPFeatureExtractor, CLIPTextModel, CLIPTokenizer, CLIPVisionModelWithProjection
+from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer, CLIPVisionModelWithProjection
 
 from diffusers import DiffusionPipeline
 from diffusers.configuration_utils import FrozenDict, deprecate
@@ -683,7 +683,7 @@ class TensorRTStableDiffusionInpaintPipeline(DiffusionPipeline):
         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 ([`CLIPFeatureExtractor`]):
+        feature_extractor ([`CLIPImageProcessor`]):
             Model that extracts features from generated images to be used as inputs for the `safety_checker`.
     """
 
@@ -697,7 +697,7 @@ def __init__(
         unet: UNet2DConditionModel,
         scheduler: DDIMScheduler,
         safety_checker: StableDiffusionSafetyChecker,
-        feature_extractor: CLIPFeatureExtractor,
+        feature_extractor: CLIPImageProcessor,
         image_encoder: CLIPVisionModelWithProjection = None,
         requires_safety_checker: bool = True,
         stages=["clip", "unet", "vae", "vae_encoder"],

examples/community/stable_diffusion_tensorrt_txt2img.py

@@ -42,7 +42,7 @@
     network_from_onnx_path,
     save_engine,
 )
-from transformers import CLIPFeatureExtractor, CLIPTextModel, CLIPTokenizer, CLIPVisionModelWithProjection
+from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer, CLIPVisionModelWithProjection
 
 from diffusers import DiffusionPipeline
 from diffusers.configuration_utils import FrozenDict, deprecate
@@ -595,7 +595,7 @@ class TensorRTStableDiffusionPipeline(DiffusionPipeline):
         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 ([`CLIPFeatureExtractor`]):
+        feature_extractor ([`CLIPImageProcessor`]):
             Model that extracts features from generated images to be used as inputs for the `safety_checker`.
     """
 
@@ -609,7 +609,7 @@ def __init__(
         unet: UNet2DConditionModel,
         scheduler: DDIMScheduler,
         safety_checker: StableDiffusionSafetyChecker,
-        feature_extractor: CLIPFeatureExtractor,
+        feature_extractor: CLIPImageProcessor,
         image_encoder: CLIPVisionModelWithProjection = None,
         requires_safety_checker: bool = True,
         stages=["clip", "unet", "vae"],

examples/research_projects/controlnet/train_controlnet_webdataset.py

@@ -43,7 +43,7 @@
 from torch.utils.data import default_collate
 from torchvision import transforms
 from tqdm.auto import tqdm
-from transformers import AutoTokenizer, DPTFeatureExtractor, DPTForDepthEstimation, PretrainedConfig
+from transformers import AutoTokenizer, DPTForDepthEstimation, DPTImageProcessor, PretrainedConfig
 from webdataset.tariterators import (
     base_plus_ext,
     tar_file_expander,
@@ -205,7 +205,7 @@ def __init__(
         pin_memory: bool = False,
         persistent_workers: bool = False,
         control_type: str = "canny",
-        feature_extractor: Optional[DPTFeatureExtractor] = None,
+        feature_extractor: Optional[DPTImageProcessor] = None,
     ):
         if not isinstance(train_shards_path_or_url, str):
             train_shards_path_or_url = [list(braceexpand(urls)) for urls in train_shards_path_or_url]
@@ -1011,7 +1011,7 @@ def main(args):
         controlnet = pre_controlnet
 
     if args.control_type == "depth":
-        feature_extractor = DPTFeatureExtractor.from_pretrained("Intel/dpt-hybrid-midas")
+        feature_extractor = DPTImageProcessor.from_pretrained("Intel/dpt-hybrid-midas")
         depth_model = DPTForDepthEstimation.from_pretrained("Intel/dpt-hybrid-midas")
         depth_model.requires_grad_(False)
     else:

examples/research_projects/gligen/demo.ipynb

@@ -45,7 +45,7 @@
     "    UniPCMultistepScheduler,\n",
     "    EulerDiscreteScheduler,\n",
     ")\n",
-    "from transformers import CLIPFeatureExtractor, CLIPTextModel, CLIPTokenizer\n",
+    "from transformers import CLIPImageProcessor, CLIPTextModel, CLIPTokenizer\n",
     "# pretrained_model_name_or_path = 'masterful/gligen-1-4-generation-text-box'\n",
     "\n",
     "pretrained_model_name_or_path = '/root/data/zhizhonghuang/checkpoints/models--masterful--gligen-1-4-generation-text-box/snapshots/d2820dc1e9ba6ca082051ce79cfd3eb468ae2c83'\n",

examples/research_projects/rdm/pipeline_rdm.py

@@ -4,7 +4,7 @@
 import torch
 from PIL import Image
 from retriever import Retriever, normalize_images, preprocess_images
-from transformers import CLIPFeatureExtractor, CLIPModel, CLIPTokenizer
+from transformers import CLIPImageProcessor, CLIPModel, CLIPTokenizer
 
 from diffusers import (
     AutoencoderKL,
@@ -47,7 +47,7 @@ class RDMPipeline(DiffusionPipeline, StableDiffusionMixin):
         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`].
-        feature_extractor ([`CLIPFeatureExtractor`]):
+        feature_extractor ([`CLIPImageProcessor`]):
             Model that extracts features from generated images to be used as inputs for the `safety_checker`.
     """
 
@@ -65,7 +65,7 @@ def __init__(
             EulerAncestralDiscreteScheduler,
             DPMSolverMultistepScheduler,
         ],
-        feature_extractor: CLIPFeatureExtractor,
+        feature_extractor: CLIPImageProcessor,
         retriever: Optional[Retriever] = None,
     ):
         super().__init__()