in the resize() function in image_transforms.py, the line 267: (#20728)

`image = to_channel_dimension_format(image, ChannelDimension.LAST)`
is redundant as this same conversion is also applied in to_pil_image().

This redundant call actually makes the training fail in rare cases.
The problem can be reproduced with the following code snippet:
```
from transformers.models.clip import CLIPFeatureExtractor
vision_processor = CLIPFeatureExtractor.from_pretrained('openai/clip-vit-large-patch14')
images = [
    torch.rand(size=(3, 2, 10), dtype=torch.float),
    torch.rand(size=(3, 10, 1), dtype=torch.float),
    torch.rand(size=(3, 1, 10), dtype=torch.float)
]
for image in images:
    processed_image = vision_processor(images=image, return_tensors="pt")['pixel_values']
    print(processed_image.shape)
    assert processed_image.shape == torch.Size([1, 3, 224, 224])
```

The last image has a height of 1 pixel.
The second call to to_channel_dimesion_format() will transpose the image, and the height
dimension is wrongly treated as the channels dimension afterwards.
Because of this, the following normalize() step will result in an
exception.

src/transformers/image_transforms.py

@@ -271,8 +271,6 @@ def resize(
     # To maintain backwards compatibility with the resizing done in previous image feature extractors, we use
     # the pillow library to resize the image and then convert back to numpy
     if not isinstance(image, PIL.Image.Image):
-        # PIL expects image to have channels last
-        image = to_channel_dimension_format(image, ChannelDimension.LAST)
         image = to_pil_image(image)
     height, width = size
     # PIL images are in the format (width, height)