/cuBERT

Fast implementation of BERT inference directly on NVIDIA (CUDA, CUBLAS) and Intel MKL

Primary LanguageC++MIT LicenseMIT

Watchers

No one’s watching this repository yet.