TransformerEngine

machine-learning

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.

How to download and setup TransformerEngine

Open terminal and run command

git clone https://github.com/NVIDIA/TransformerEngine.git

git clone is used to create a copy or clone of TransformerEngine repositories. You pass git clone a repository URL.
it supports a few different network protocols and corresponding URL formats.

Also you may download zip file with TransformerEngine https://github.com/NVIDIA/TransformerEngine/archive/master.zip

Or simply clone TransformerEngine with SSH

[email protected]:NVIDIA/TransformerEngine.git

If you have some problems with TransformerEngine

You may open issue on TransformerEngine support forum (system) here: https://github.com/NVIDIA/TransformerEngine/issues

Similar to TransformerEngine repositories

Here you may see TransformerEngine alternatives and analogs

tensorflow keras scikit-learn TensorFlow-Examples pytorch face_recognition CNTK data-science-ipython-notebooks Qix handong1587.github.io telegram-list netdata mlcourse.ai stats Winds machine-learning-curriculum caffe tesseract machine-learning-for-software-engineers awesome-deep-learning-papers incubator-mxnet lectures cs-video-courses julia Screenshot-to-code spaCy cheatsheets-ai awesome-deep-learning python-machine-learning-book WaveFunctionCollapse