TransformerEngine

0

Описание

A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper and Ada GPUs, to provide better performance with lower memory utilization in both training and inference.

Языки

Python

  • CMake
  • Cuda
  • C++
  • C
  • Shell
Сообщить о нарушении