Transformers

Implementing Self-Attention from Scratch in PyTorch

Implementing Self-Attention from Scratch in PyTorch

Another medium article: The Easiest Way to Understand Self-Attention with Code.

Attention Is All You Need

Introduces the Transformer, a novel neural network architecture based solely on attention mechanisms for sequence transduction, improving machine translation quality, training speed, and parallelization over recurrent and convolutional models.

proceedings.neurips.cc

Attention is all you need paper

Transformers Explained Visually (Part 1): Overview of Functionality | Towards Data Science

towardsdatascience.com

Transformers Explained Visually I

Just a moment...

Mriganka Nath mrinath.medium.com

Patchifying Images for Computer Vision

arxiv.org

Vision TransFormers original paper

How to Use the Segment Anything Model (SAM)

Piotr Skalski blog.roboflow.com

Transformers from Scratch

links:

1. Training Transformers from Scratch: A Deep Dive Guide | by Devang Vashistha | in Data Science Collective - Freedium

2. https://differ.blog/p/here-s-how-you-can-build-and-train-gpt-2-from-scratch-using-pytorch-ace4ba

google.com

SAM model

Ideas related to this collection