/attention_with_linear_biases

Code for the ALiBi method for transformer language models (ICLR 2022)

Primary LanguagePythonMIT LicenseMIT

No issues in this repository yet.