Stars
2
stars
written in C++
Clear filter
Fast implementation of BERT inference directly on NVIDIA (CUDA, CUBLAS) and Intel MKL