Update New Features and Update Documents in V0.2.6

[Lupin1998]

Updating features:

1. Support new self-supervised method BEiT with ViT-Base on ImageNet-1K, and fix bugs of CAE, MaskFeat, and SimMIM in Dataset, Model, and Head. Note that we added HOG feature implementation borrowed from the original repo for MaskFeat.
2. Support new backbone architecture DeiT-3 and provide configs.
3. Update pre-training and fine-tuning config files, and documents for the relevant masked image modeling (MIM) methods (BEiT, MaskFeat, CAE, and A2MIM).
4. Support Grad-CAM visualization tools vis_cam.py of supported architectures.

Updating documents:

1. Update the template and add the latest paper lists of mixup and MIM methods in Awesome Mixups and Awesome MIM.
2. Update documents of tools.

[Lupin1998]

Fixed bugs and updated features for MIM methods in OpenMixup V0.2.6.

[Lupin1998]

Update new features and documents for OpenMixup V0.2.6 (issue #24).