Module vit 

Vision Transformer (ViT) implementation.

Vision Transformer applies transformer architecture to image classification by splitting images into patches and processing them as a sequence.

Key characteristics:

• Image patches as sequence tokens
• Self-attention between patches
• Position embeddings
• CLS token for classification
• Layer normalization

References:

• ViT Paper
• Model Card

Structs

Config

Embeddings

Encoder

Model