diffusers 0.33.1__py3-none-any.whl → 0.34.0__py3-none-any.whl

diffusers/pipelines/kandinsky2_2/pipeline_kandinsky2_2_combined.py

@@ -1,4 +1,4 @@
-# Copyright 2024 The HuggingFace Team. All rights reserved.
+# Copyright 2025 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.
@@ -179,7 +179,7 @@ class KandinskyV22CombinedPipeline(DiffusionPipeline):
     def enable_xformers_memory_efficient_attention(self, attention_op: Optional[Callable] = None):
         self.decoder_pipe.enable_xformers_memory_efficient_attention(attention_op)
 
-    def enable_sequential_cpu_offload(self, gpu_id: Optional[int] = None, device: Union[torch.device, str] = "cuda"):
+    def enable_sequential_cpu_offload(self, gpu_id: Optional[int] = None, device: Union[torch.device, str] = None):
         r"""
         Offloads all models to CPU using accelerate, significantly reducing memory usage. When called, unet,
         text_encoder, vae and safety checker have their state dicts saved to CPU and then are moved to a
@@ -242,20 +242,20 @@ class KandinskyV22CombinedPipeline(DiffusionPipeline):
             width (`int`, *optional*, defaults to 512):
                 The width in pixels of the generated image.
             prior_guidance_scale (`float`, *optional*, defaults to 4.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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.
             prior_num_inference_steps (`int`, *optional*, defaults to 100):
                 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 4.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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.
             generator (`torch.Generator` or `List[torch.Generator]`, *optional*):
                 One or a list of [torch generator(s)](https://pytorch.org/docs/stable/generated/torch.Generator.html)
                 to make generation deterministic.
@@ -407,7 +407,7 @@ class KandinskyV22Img2ImgCombinedPipeline(DiffusionPipeline):
     def enable_xformers_memory_efficient_attention(self, attention_op: Optional[Callable] = None):
         self.decoder_pipe.enable_xformers_memory_efficient_attention(attention_op)
 
-    def enable_model_cpu_offload(self, gpu_id: Optional[int] = None, device: Union[torch.device, str] = "cuda"):
+    def enable_model_cpu_offload(self, gpu_id: Optional[int] = None, device: Union[torch.device, str] = None):
         r"""
         Offloads all models to CPU using accelerate, reducing memory usage with a low impact on performance. Compared
         to `enable_sequential_cpu_offload`, this method moves one whole model at a time to the GPU when its `forward`
@@ -417,7 +417,7 @@ class KandinskyV22Img2ImgCombinedPipeline(DiffusionPipeline):
         self.prior_pipe.enable_model_cpu_offload(gpu_id=gpu_id, device=device)
         self.decoder_pipe.enable_model_cpu_offload(gpu_id=gpu_id, device=device)
 
-    def enable_sequential_cpu_offload(self, gpu_id: Optional[int] = None, device: Union[torch.device, str] = "cuda"):
+    def enable_sequential_cpu_offload(self, gpu_id: Optional[int] = None, device: Union[torch.device, str] = None):
         r"""
         Offloads all models to CPU using accelerate, significantly reducing memory usage. When called, unet,
         text_encoder, vae and safety checker have their state dicts saved to CPU and then are moved to a
@@ -479,11 +479,11 @@ class KandinskyV22Img2ImgCombinedPipeline(DiffusionPipeline):
             num_images_per_prompt (`int`, *optional*, defaults to 1):
                 The number of images to generate per prompt.
             guidance_scale (`float`, *optional*, defaults to 4.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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.
             strength (`float`, *optional*, defaults to 0.3):
                 Conceptually, indicates how much to transform the reference `image`. Must be between 0 and 1. `image`
                 will be used as a starting point, adding more noise to it the larger the `strength`. The number of
@@ -498,11 +498,11 @@ class KandinskyV22Img2ImgCombinedPipeline(DiffusionPipeline):
             width (`int`, *optional*, defaults to 512):
                 The width in pixels of the generated image.
             prior_guidance_scale (`float`, *optional*, defaults to 4.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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.
             prior_num_inference_steps (`int`, *optional*, defaults to 100):
                 The number of denoising steps. More denoising steps usually lead to a higher quality image at the
                 expense of slower inference.
@@ -656,7 +656,7 @@ class KandinskyV22InpaintCombinedPipeline(DiffusionPipeline):
     def enable_xformers_memory_efficient_attention(self, attention_op: Optional[Callable] = None):
         self.decoder_pipe.enable_xformers_memory_efficient_attention(attention_op)
 
-    def enable_sequential_cpu_offload(self, gpu_id: Optional[int] = None, device: Union[torch.device, str] = "cuda"):
+    def enable_sequential_cpu_offload(self, gpu_id: Optional[int] = None, device: Union[torch.device, str] = None):
         r"""
         Offloads all models to CPU using accelerate, significantly reducing memory usage. When called, unet,
         text_encoder, vae and safety checker have their state dicts saved to CPU and then are moved to a
@@ -722,11 +722,11 @@ class KandinskyV22InpaintCombinedPipeline(DiffusionPipeline):
             num_images_per_prompt (`int`, *optional*, defaults to 1):
                 The number of images to generate per prompt.
             guidance_scale (`float`, *optional*, defaults to 4.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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.
             num_inference_steps (`int`, *optional*, defaults to 100):
                 The number of denoising steps. More denoising steps usually lead to a higher quality image at the
                 expense of slower inference.
@@ -735,11 +735,11 @@ class KandinskyV22InpaintCombinedPipeline(DiffusionPipeline):
             width (`int`, *optional*, defaults to 512):
                 The width in pixels of the generated image.
             prior_guidance_scale (`float`, *optional*, defaults to 4.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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.
             prior_num_inference_steps (`int`, *optional*, defaults to 100):
                 The number of denoising steps. More denoising steps usually lead to a higher quality image at the
                 expense of slower inference.

diffusers/pipelines/kandinsky2_2/pipeline_kandinsky2_2_controlnet.py

@@ -1,4 +1,4 @@
-# Copyright 2024 The HuggingFace Team. All rights reserved.
+# Copyright 2025 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.
@@ -198,11 +198,11 @@ class KandinskyV22ControlnetPipeline(DiffusionPipeline):
                 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 4.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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.
             num_images_per_prompt (`int`, *optional*, defaults to 1):
                 The number of images to generate per prompt.
             generator (`torch.Generator` or `List[torch.Generator]`, *optional*):

diffusers/pipelines/kandinsky2_2/pipeline_kandinsky2_2_controlnet_img2img.py

@@ -1,4 +1,4 @@
-# Copyright 2024 The HuggingFace Team. All rights reserved.
+# Copyright 2025 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.
@@ -14,11 +14,10 @@
 
 from typing import Callable, List, Optional, Union
 
-import numpy as np
 import PIL.Image
 import torch
-from PIL import Image
 
+from ...image_processor import VaeImageProcessor
 from ...models import UNet2DConditionModel, VQModel
 from ...schedulers import DDPMScheduler
 from ...utils import (
@@ -105,27 +104,6 @@ EXAMPLE_DOC_STRING = """
 """
 
 
-# Copied from diffusers.pipelines.kandinsky2_2.pipeline_kandinsky2_2.downscale_height_and_width
-def downscale_height_and_width(height, width, scale_factor=8):
-    new_height = height // scale_factor**2
-    if height % scale_factor**2 != 0:
-        new_height += 1
-    new_width = width // scale_factor**2
-    if width % scale_factor**2 != 0:
-        new_width += 1
-    return new_height * scale_factor, new_width * scale_factor
-
-
-# Copied from diffusers.pipelines.kandinsky.pipeline_kandinsky_img2img.prepare_image
-def prepare_image(pil_image, w=512, h=512):
-    pil_image = pil_image.resize((w, h), resample=Image.BICUBIC, reducing_gap=1)
-    arr = np.array(pil_image.convert("RGB"))
-    arr = arr.astype(np.float32) / 127.5 - 1
-    arr = np.transpose(arr, [2, 0, 1])
-    image = torch.from_numpy(arr).unsqueeze(0)
-    return image
-
-
 class KandinskyV22ControlnetImg2ImgPipeline(DiffusionPipeline):
     """
     Pipeline for image-to-image generation using Kandinsky
@@ -157,7 +135,14 @@ class KandinskyV22ControlnetImg2ImgPipeline(DiffusionPipeline):
             scheduler=scheduler,
             movq=movq,
         )
-        self.movq_scale_factor = 2 ** (len(self.movq.config.block_out_channels) - 1)
+        movq_scale_factor = 2 ** (len(self.movq.config.block_out_channels) - 1) if getattr(self, "movq", None) else 8
+        movq_latent_channels = self.movq.config.latent_channels if getattr(self, "movq", None) else 4
+        self.image_processor = VaeImageProcessor(
+            vae_scale_factor=movq_scale_factor,
+            vae_latent_channels=movq_latent_channels,
+            resample="bicubic",
+            reducing_gap=1,
+        )
 
     # Copied from diffusers.pipelines.kandinsky.pipeline_kandinsky_img2img.KandinskyImg2ImgPipeline.get_timesteps
     def get_timesteps(self, num_inference_steps, strength, device):
@@ -259,11 +244,11 @@ class KandinskyV22ControlnetImg2ImgPipeline(DiffusionPipeline):
                 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 4.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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.
             num_images_per_prompt (`int`, *optional*, defaults to 1):
                 The number of images to generate per prompt.
             generator (`torch.Generator` or `List[torch.Generator]`, *optional*):
@@ -316,7 +301,7 @@ class KandinskyV22ControlnetImg2ImgPipeline(DiffusionPipeline):
                 f"Input is in incorrect format: {[type(i) for i in image]}. Currently, we only support  PIL image and pytorch tensor"
             )
 
-        image = torch.cat([prepare_image(i, width, height) for i in image], dim=0)
+        image = torch.cat([self.image_processor.preprocess(i, width, height) for i in image], dim=0)
         image = image.to(dtype=image_embeds.dtype, device=device)
 
         latents = self.movq.encode(image)["latents"]
@@ -324,7 +309,6 @@ class KandinskyV22ControlnetImg2ImgPipeline(DiffusionPipeline):
         self.scheduler.set_timesteps(num_inference_steps, device=device)
         timesteps, num_inference_steps = self.get_timesteps(num_inference_steps, strength, device)
         latent_timestep = timesteps[:1].repeat(batch_size * num_images_per_prompt)
-        height, width = downscale_height_and_width(height, width, self.movq_scale_factor)
         latents = self.prepare_latents(
             latents, latent_timestep, batch_size, num_images_per_prompt, image_embeds.dtype, device, generator
         )
@@ -379,13 +363,7 @@ class KandinskyV22ControlnetImg2ImgPipeline(DiffusionPipeline):
         if output_type not in ["pt", "np", "pil"]:
             raise ValueError(f"Only the output types `pt`, `pil` and `np` are supported not output_type={output_type}")
 
-        if output_type in ["np", "pil"]:
-            image = image * 0.5 + 0.5
-            image = image.clamp(0, 1)
-            image = image.cpu().permute(0, 2, 3, 1).float().numpy()
-
-        if output_type == "pil":
-            image = self.numpy_to_pil(image)
+        image = self.image_processor.postprocess(image, output_type)
 
         if not return_dict:
             return (image,)

diffusers/pipelines/kandinsky2_2/pipeline_kandinsky2_2_img2img.py

@@ -1,4 +1,4 @@
-# Copyright 2024 The HuggingFace Team. All rights reserved.
+# Copyright 2025 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.
@@ -14,11 +14,10 @@
 
 from typing import Callable, Dict, List, Optional, Union
 
-import numpy as np
 import PIL.Image
 import torch
-from PIL import Image
 
+from ...image_processor import VaeImageProcessor
 from ...models import UNet2DConditionModel, VQModel
 from ...schedulers import DDPMScheduler
 from ...utils import deprecate, is_torch_xla_available, logging
@@ -76,27 +75,6 @@ EXAMPLE_DOC_STRING = """
 """
 
 
-# Copied from diffusers.pipelines.kandinsky2_2.pipeline_kandinsky2_2.downscale_height_and_width
-def downscale_height_and_width(height, width, scale_factor=8):
-    new_height = height // scale_factor**2
-    if height % scale_factor**2 != 0:
-        new_height += 1
-    new_width = width // scale_factor**2
-    if width % scale_factor**2 != 0:
-        new_width += 1
-    return new_height * scale_factor, new_width * scale_factor
-
-
-# Copied from diffusers.pipelines.kandinsky.pipeline_kandinsky_img2img.prepare_image
-def prepare_image(pil_image, w=512, h=512):
-    pil_image = pil_image.resize((w, h), resample=Image.BICUBIC, reducing_gap=1)
-    arr = np.array(pil_image.convert("RGB"))
-    arr = arr.astype(np.float32) / 127.5 - 1
-    arr = np.transpose(arr, [2, 0, 1])
-    image = torch.from_numpy(arr).unsqueeze(0)
-    return image
-
-
 class KandinskyV22Img2ImgPipeline(DiffusionPipeline):
     """
     Pipeline for image-to-image generation using Kandinsky
@@ -129,7 +107,14 @@ class KandinskyV22Img2ImgPipeline(DiffusionPipeline):
             scheduler=scheduler,
             movq=movq,
         )
-        self.movq_scale_factor = 2 ** (len(self.movq.config.block_out_channels) - 1)
+        movq_scale_factor = 2 ** (len(self.movq.config.block_out_channels) - 1) if getattr(self, "movq", None) else 8
+        movq_latent_channels = self.movq.config.latent_channels if getattr(self, "movq", None) else 4
+        self.image_processor = VaeImageProcessor(
+            vae_scale_factor=movq_scale_factor,
+            vae_latent_channels=movq_latent_channels,
+            resample="bicubic",
+            reducing_gap=1,
+        )
 
     # Copied from diffusers.pipelines.kandinsky.pipeline_kandinsky_img2img.KandinskyImg2ImgPipeline.get_timesteps
     def get_timesteps(self, num_inference_steps, strength, device):
@@ -240,11 +225,11 @@ class KandinskyV22Img2ImgPipeline(DiffusionPipeline):
                 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 4.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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.
             num_images_per_prompt (`int`, *optional*, defaults to 1):
                 The number of images to generate per prompt.
             generator (`torch.Generator` or `List[torch.Generator]`, *optional*):
@@ -319,7 +304,7 @@ class KandinskyV22Img2ImgPipeline(DiffusionPipeline):
                 f"Input is in incorrect format: {[type(i) for i in image]}. Currently, we only support  PIL image and pytorch tensor"
             )
 
-        image = torch.cat([prepare_image(i, width, height) for i in image], dim=0)
+        image = torch.cat([self.image_processor.preprocess(i, width, height) for i in image], dim=0)
         image = image.to(dtype=image_embeds.dtype, device=device)
 
         latents = self.movq.encode(image)["latents"]
@@ -327,7 +312,6 @@ class KandinskyV22Img2ImgPipeline(DiffusionPipeline):
         self.scheduler.set_timesteps(num_inference_steps, device=device)
         timesteps, num_inference_steps = self.get_timesteps(num_inference_steps, strength, device)
         latent_timestep = timesteps[:1].repeat(batch_size * num_images_per_prompt)
-        height, width = downscale_height_and_width(height, width, self.movq_scale_factor)
         latents = self.prepare_latents(
             latents, latent_timestep, batch_size, num_images_per_prompt, image_embeds.dtype, device, generator
         )
@@ -383,21 +367,9 @@ class KandinskyV22Img2ImgPipeline(DiffusionPipeline):
             if XLA_AVAILABLE:
                 xm.mark_step()
 
-        if output_type not in ["pt", "np", "pil", "latent"]:
-            raise ValueError(
-                f"Only the output types `pt`, `pil` ,`np` and `latent` are supported not output_type={output_type}"
-            )
-
         if not output_type == "latent":
-            # post-processing
             image = self.movq.decode(latents, force_not_quantize=True)["sample"]
-            if output_type in ["np", "pil"]:
-                image = image * 0.5 + 0.5
-                image = image.clamp(0, 1)
-                image = image.cpu().permute(0, 2, 3, 1).float().numpy()
-
-            if output_type == "pil":
-                image = self.numpy_to_pil(image)
+            image = self.image_processor.postprocess(image, output_type)
         else:
             image = latents
 

diffusers/pipelines/kandinsky2_2/pipeline_kandinsky2_2_inpainting.py

@@ -1,4 +1,4 @@
-# Copyright 2024 The HuggingFace Team. All rights reserved.
+# Copyright 2025 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.
@@ -343,11 +343,11 @@ class KandinskyV22InpaintPipeline(DiffusionPipeline):
                 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 4.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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.
             num_images_per_prompt (`int`, *optional*, defaults to 1):
                 The number of images to generate per prompt.
             generator (`torch.Generator` or `List[torch.Generator]`, *optional*):
@@ -386,7 +386,7 @@ class KandinskyV22InpaintPipeline(DiffusionPipeline):
                 "As of diffusers==0.19.0 this behavior has been inverted. Now white pixels are repainted and black pixels are preserved. "
                 "This way, Kandinsky's masking behavior is aligned with Stable Diffusion. "
                 "THIS means that you HAVE to invert the input mask to have the same behavior as before as explained in https://github.com/huggingface/diffusers/pull/4207. "
-                "This warning will be surpressed after the first inference call and will be removed in diffusers>0.23.0"
+                "This warning will be suppressed after the first inference call and will be removed in diffusers>0.23.0"
             )
             self._warn_has_been_called = True
 

diffusers/pipelines/kandinsky2_2/pipeline_kandinsky2_2_prior.py

@@ -179,11 +179,11 @@ class KandinskyV22PriorPipeline(DiffusionPipeline):
                 The prompt not to guide the image generation. Ignored when not using guidance (i.e., ignored if
                 `guidance_scale` is less than `1`).
             guidance_scale (`float`, *optional*, defaults to 4.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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.
 
         Examples:
 
@@ -414,11 +414,11 @@ class KandinskyV22PriorPipeline(DiffusionPipeline):
                 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`.
             guidance_scale (`float`, *optional*, defaults to 4.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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.
             output_type (`str`, *optional*, defaults to `"pt"`):
                 The output format of the generate image. Choose between: `"np"` (`np.array`) or `"pt"`
                 (`torch.Tensor`).

diffusers/pipelines/kandinsky2_2/pipeline_kandinsky2_2_prior_emb2emb.py

@@ -203,11 +203,11 @@ class KandinskyV22PriorEmb2EmbPipeline(DiffusionPipeline):
                 The prompt not to guide the image generation. Ignored when not using guidance (i.e., ignored if
                 `guidance_scale` is less than `1`).
             guidance_scale (`float`, *optional*, defaults to 4.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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.
 
         Examples:
 
@@ -441,11 +441,11 @@ class KandinskyV22PriorEmb2EmbPipeline(DiffusionPipeline):
                 One or a list of [torch generator(s)](https://pytorch.org/docs/stable/generated/torch.Generator.html)
                 to make generation deterministic.
             guidance_scale (`float`, *optional*, defaults to 4.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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.
             output_type (`str`, *optional*, defaults to `"pt"`):
                 The output format of the generate image. Choose between: `"np"` (`np.array`) or `"pt"`
                 (`torch.Tensor`).

diffusers/pipelines/kandinsky3/pipeline_kandinsky3.py

@@ -368,11 +368,11 @@ class Kandinsky3Pipeline(DiffusionPipeline, StableDiffusionLoraLoaderMixin):
                 Custom timesteps to use for the denoising process. If not defined, equal spaced `num_inference_steps`
                 timesteps are used. Must be in descending order.
             guidance_scale (`float`, *optional*, defaults to 3.0):
-                Guidance scale as defined in [Classifier-Free Diffusion Guidance](https://arxiv.org/abs/2207.12598).
-                `guidance_scale` is defined as `w` of equation 2. of [Imagen
-                Paper](https://arxiv.org/pdf/2205.11487.pdf). Guidance scale is enabled by setting `guidance_scale >
-                1`. Higher guidance scale encourages to generate images that are closely linked to the text `prompt`,
-                usually at the expense of lower image quality.
+                Guidance scale as defined in [Classifier-Free Diffusion
+                Guidance](https://huggingface.co/papers/2207.12598). `guidance_scale` is defined as `w` of equation 2.
+                of [Imagen Paper](https://huggingface.co/papers/2205.11487). 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
@@ -384,8 +384,8 @@ class Kandinsky3Pipeline(DiffusionPipeline, StableDiffusionLoraLoaderMixin):
             width (`int`, *optional*, defaults to self.unet.config.sample_size):
                 The width in pixels of the generated image.
             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.
+                Corresponds to parameter eta (η) in the DDIM paper: https://huggingface.co/papers/2010.02502. Only
+                applies to [`schedulers.DDIMScheduler`], will be ignored for others.
             generator (`torch.Generator` or `List[torch.Generator]`, *optional*):
                 One or a list of [torch generator(s)](https://pytorch.org/docs/stable/generated/torch.Generator.html)
                 to make generation deterministic.