Skip to content

Latest commit

 

History

History
7 lines (4 loc) · 383 Bytes

README.md

File metadata and controls

7 lines (4 loc) · 383 Bytes

hedgehogT

Implementation of hedgehog transformer which is a Linear Attention Mechanism inspired by the concept described in "The Hedgehog & the Porcupine: Expressive Linear Attentions with Softmax Mimicry."

https://arxiv.org/pdf/2402.04347.pdf

Screenshot from 2024-02-09 16-00-03