Add Support for Vision Transformer Models (ViT/Uformer) and Extend for MRI Reconstruction

[georgeyiasemis]

Description
Implement Vision Transformer (ViT) and Uformer models in the codebase to support tasks related to vision-based deep learning. Extend these models specifically for MRI reconstruction.

Tasks

• Implement standard ViT and Uformer architectures.
• Modify the models to handle MRI data formats (e.g., k-space, image domain).
• Provide tests for both vision tasks and MRI reconstruction.