/galerkin-transformer

[NeurIPS 2021] Galerkin Transformer: a linear attention without softmax

Primary LanguagePythonMIT LicenseMIT

Stargazers

No one’s star this repository yet.