markriedl / transformer-walkthrough

A walkthrough of transformer architecture code

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

A walkthrough of transformer architecture code

Open In Colab

The notebook walks through a single forward pass of the Transformer architecture in pytorch. It is meant for illustration and educational purposes only. The walkthrough explains every stage of the architecture accompanied by a detailed computation graph.

Transformer Computation Graph

About

A walkthrough of transformer architecture code

License:MIT License


Languages

Language:Jupyter Notebook 100.0%