[zucchini-nlp]

What does this PR do?

This PR allows users to pass the device to processor call and get all inputs in the requested device. Currently the device arg is used only in fast image/video processors. But users get confused by the following case

inputs = processor(images, text, device='cuda', return_tensors='pt')

# Before the PR
print(inputs.input_ids.device) # "cpu"
print(inputs.pixel_values.device) # "cuda"

# After the PR
print(inputs.input_ids.device) # "cuda"
print(inputs.pixel_values.device) # "cuda

Now we will allow either to pass device and get all inputs initialized already on device, or to move to device manually by inputs.to(device)

Fixes #42722

[github-actions]

View the CircleCI Test Summary for this PR:

https://huggingface.co/spaces/transformers-community/circle-ci-viz?pr=43249&sha=703f75

[ArthurZucker]

Closing this WIP: the image-processor backend refactor removed image_processing_utils_fast.py and the *_fast processors it edits, so it can't be rebased as it is. Moving outputs to device in __call__ is still a good idea if we want to pick it up again on the new backend layout.

[zucchini-nlp]

wait dont close, i will forget it ever existed otherwise 😆 This is my reminder