Note: the session will be recorded for registrants who cannot attend at the scheduled time.

The transformer architecture is the basis for ChatGPT and other large language models. Transformers can be used with computer vision tasks by serializing images into sequences of patches. This architecture is arguably superior to traditional convolutional neural networks (CNNs) in several ways. We will build and test a Vision Transformer (ViT) from scratch in Python and demonstrate powerful, existing ViT Python libraries.