offer way for extractor to return latents without detaching them

slightly more clear of einops rearrange for cls token, for lucidrains…

…#224

follow @arquolo feedback and advice for MaxViT

fix hidden dimension in MaxViT thanks to @arquolo

be able to accept a reference to a layer within the model for forward…

… hooking and extracting the embedding output, for regionvit to work with extractor

make extractor flexible for layers that output multiple tensors, show…

… CrossViT example

add simple vit, from https://arxiv.org/abs/2205.01580

add simple vit, from https://arxiv.org/abs/2205.01580

add simple vit, from https://arxiv.org/abs/2205.01580

add EsViT, by popular request, an alternative to Dino that is compati…

…ble with efficient ViTs with accounting for regional self-supervised loss