# Vision Transformer (ViT) Configurations

## ViT Tiny
- **Model Name:** ViT Tiny
- **Patch Size:** 16x16 pixels
- **Embedding Dimension (d):** Typically around 192
- **Number of Attention Heads (h):** 3 or 6
- **Number of Transformer Blocks:** Around 6
- **Total Parameters:** Fewer parameters compared to larger variants

## ViT Base
- **Model Name:** ViT Base
- **Patch Size:** 16x16 pixels
- **Embedding Dimension (d):** 768 or 1024
- **Number of Attention Heads (h):** 12 or 16
- **Number of Transformer Blocks:** Around 12
- **Total Parameters:** Moderate number of parameters

## ViT Large
- **Model Name:** ViT Large
- **Patch Size:** 16x16 pixels
- **Embedding Dimension (d):** 1024 or 1280
- **Number of Attention Heads (h):** 16
- **Number of Transformer Blocks:** Around 24
- **Total Parameters:** Large number of parameters