NVIDIA/TransformerEngine

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.

PythonCudaC++COtherpythonmachine-learningdeep-learninggpucudapytorchjaxfp8
This is stars and forks stats for /NVIDIA/TransformerEngine repository. As of 03 May, 2024 this repository has 970 stars and 124 forks.

Transformer Engine Quickstart | Installation | User Guide | Examples | Model Support | Integrations | Release notes Latest News [04/2023] Benchmarking Large Language Models on NVIDIA H100 GPUs with CoreWeave (Part 1) What is Transformer Engine? Transformer Engine (TE) is a library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit floating point (FP8) precision on Hopper GPUs, to provide better performance with lower memory utilization in both training and inference. TE provides...
Read on GithubGithub Stats Page
repotechsstarsweeklyforksweekly
gihannagalindez/boostdiff_inferenceCythonRPython10010
Trinity-Developers-Club/20DOOS_APP_DEVDartJavaC++80420
maxiee/dighubDartC++CMake53020
beardedio/terrariaDockerfilePythonShell890520
findmypast/hex_webElixirHTMLCSS1000
roovo/obsidian-card-boardElmTypeScriptCSS3190140
yaelahaiz/hacktober-project2PHPC++Python00160
reanimate/reanimateHaskellElmOther1.1k0570
TheRetroSpecter/VsRetro-Psych-PublicHaxeCLua150190
weiweihuanghuang/fragment-monoHTMLPythonMakefile626+480