Implementation of Griffin from the paper: "Griffin: Mixing Gated Linear Recurrences with Local Attention for Efficient Language Models"
Primary LanguagePythonMIT LicenseMIT