code quality updates

Author: ishan-modi

src/diffusers/image_processor.py

@@ -499,7 +499,11 @@ def resize(
             raise ValueError(f"Only PIL image input is supported for resize_mode {resize_mode}")
         if isinstance(image, PIL.Image.Image):
             if resize_mode == "default":
-                image = image.resize((width, height), resample=PIL_INTERPOLATION[self.config.resample], reducing_gap=self.config.reducing_gap)
+                image = image.resize(
+                    (width, height),
+                    resample=PIL_INTERPOLATION[self.config.resample],
+                    reducing_gap=self.config.reducing_gap,
+                )
             elif resize_mode == "fill":
                 image = self._resize_and_fill(image, width, height)
             elif resize_mode == "crop":

src/diffusers/pipelines/kandinsky2_2/pipeline_kandinsky2_2_controlnet_img2img.py

@@ -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 (
@@ -27,7 +26,7 @@
 )
 from ...utils.torch_utils import randn_tensor
 from ..pipeline_utils import DiffusionPipeline, ImagePipelineOutput
-from ...image_processor import VaeImageProcessor
+
 
 if is_torch_xla_available():
     import torch_xla.core.xla_model as xm
@@ -136,12 +135,12 @@ def __init__(
             scheduler=scheduler,
             movq=movq,
         )
-        movq_scale_factor = 2 ** (len(self.movq.config.block_out_channels) - 1) 
+        movq_scale_factor = 2 ** (len(self.movq.config.block_out_channels) - 1)
         self.image_processor = VaeImageProcessor(
-            vae_scale_factor = movq_scale_factor,
-            vae_latent_channels = self.movq.config.latent_channels,
-            resample = "bicubic",
-            reducing_gap = 1,
+            vae_scale_factor=movq_scale_factor,
+            vae_latent_channels=self.movq.config.latent_channels,
+            resample="bicubic",
+            reducing_gap=1,
         )
 
     # Copied from diffusers.pipelines.kandinsky.pipeline_kandinsky_img2img.KandinskyImg2ImgPipeline.get_timesteps

src/diffusers/pipelines/kandinsky2_2/pipeline_kandinsky2_2_img2img.py

@@ -14,17 +14,16 @@
 
 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
 from ...utils.torch_utils import randn_tensor
 from ..pipeline_utils import DiffusionPipeline, ImagePipelineOutput
-from ...image_processor import VaeImageProcessor
+
 
 if is_torch_xla_available():
     import torch_xla.core.xla_model as xm
@@ -108,12 +107,12 @@ def __init__(
             scheduler=scheduler,
             movq=movq,
         )
-        movq_scale_factor = 2 ** (len(self.movq.config.block_out_channels) - 1) 
+        movq_scale_factor = 2 ** (len(self.movq.config.block_out_channels) - 1)
         self.image_processor = VaeImageProcessor(
-            vae_scale_factor = movq_scale_factor,
-            vae_latent_channels = self.movq.config.latent_channels,
-            resample = "bicubic",
-            reducing_gap = 1,
+            vae_scale_factor=movq_scale_factor,
+            vae_latent_channels=self.movq.config.latent_channels,
+            resample="bicubic",
+            reducing_gap=1,
         )
 
     # Copied from diffusers.pipelines.kandinsky.pipeline_kandinsky_img2img.KandinskyImg2ImgPipeline.get_timesteps
@@ -372,7 +371,7 @@ def __call__(
         else:
             image = latents
         image = self.image_processor.postprocess(image, output_type=output_type)
-        
+
         # Offload all models
         self.maybe_free_model_hooks()
 

src/diffusers/pipelines/kandinsky3/pipeline_kandinsky3_img2img.py

@@ -1,12 +1,12 @@
 import inspect
 from typing import Callable, Dict, List, Optional, Union
 
-import numpy as np
 import PIL
 import PIL.Image
 import torch
 from transformers import T5EncoderModel, T5Tokenizer
 
+from ...image_processor import VaeImageProcessor
 from ...loaders import StableDiffusionLoraLoaderMixin
 from ...models import Kandinsky3UNet, VQModel
 from ...schedulers import DDPMScheduler
@@ -18,7 +18,7 @@
 )
 from ...utils.torch_utils import randn_tensor
 from ..pipeline_utils import DiffusionPipeline, ImagePipelineOutput
-from ...image_processor import VaeImageProcessor
+
 
 if is_torch_xla_available():
     import torch_xla.core.xla_model as xm
@@ -76,12 +76,12 @@ def __init__(
         self.register_modules(
             tokenizer=tokenizer, text_encoder=text_encoder, unet=unet, scheduler=scheduler, movq=movq
         )
-        movq_scale_factor = 2 ** (len(self.movq.config.block_out_channels) - 1) 
+        movq_scale_factor = 2 ** (len(self.movq.config.block_out_channels) - 1)
         self.image_processor = VaeImageProcessor(
-            vae_scale_factor = movq_scale_factor,
-            vae_latent_channels = self.movq.config.latent_channels,
-            resample = "bicubic",
-            reducing_gap = 1,
+            vae_scale_factor=movq_scale_factor,
+            vae_latent_channels=self.movq.config.latent_channels,
+            resample="bicubic",
+            reducing_gap=1,
         )
 
     def get_timesteps(self, num_inference_steps, strength, device):