Fast implementation of BERT inference directly on NVIDIA (CUDA, CUBLAS) and Intel MKL
Repository from Github https://github.comMeteorix/cuBERTRepository from Github https://github.comMeteorix/cuBERT