Fast implementation of BERT inference directly on NVIDIA (CUDA, CUBLAS) and Intel MKL
Geek Repo:Geek Repo
Github PK Tool:Github PK Tool