Fixed device selection

[ambitious-octopus]

Pull Request Description:

This PR fixes issue #1189. This PR also allows better import of the set_working_device function, facilitating more effective DeviceManager singleton management.

Example Usage:

from model_compression_toolkit.core.pytorch import set_working_device

# Set the working device to 'GPU' or 'CPU'
set_working_device('GPU')

With this change, the set_working_device function can be easily imported and used to configure the DeviceManager's device context. This also allows a quantization to be tested on a specific device.

cc: @Idan-BenAmi

Checklist before requesting a review:

• [ ] I set the appropriate labels on the pull request.
• [ ] I have added/updated the release note draft (if necessary).
• [ ] I have updated the documentation to reflect my changes (if necessary).
• [X] All function and files are well documented.
• [X] All function and classes have type hints.
• [X] There is a licenses in all file.
• [X] The function and variable names are informative.
• [X] I have checked for code duplications.
• [ ] I have added new unittest (if necessary).

[Idan-BenAmi]

Hi @ambitious-octopus, we've found the root cause of this issue (model.to() error), and we think https://github.com/sony/model_optimization/pull/1204 is a better solution.