A tiny Vision Transformer with register tokens, patch size 8, trained on ImageNet-21k with augmentation and regularization (AugReg).
No lifecycle dates recorded.
No resource links recorded.