The main processing blocks of ViT. Excludes things like patch embedding and classificaton layer. Args: width: size of the feature dim

Masked Autoencoder LightningModule. Args: image_size: (height, width) of the input images. patch_size: size of the image patches.

Masked Autoencoder for visual representation learning. Args: image_size: (height, width) of the input images. patch_size: side le

Frozen MAE encoder with trainable linear readout to class labels https://lightning.ai/docs/pytorch/stable/advanced/transfer_learning.html