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.

Date Created 2022-09-20 (2 years ago)

Commits 914 (last one 8 hours ago)

Stargazers 2,245 (17 this week)

Watchers 35 (-2 this week)

Forks 374

License apache-2.0

Ranking

RepositoryStats indexes 623,832 repositories, of these NVIDIA/TransformerEngine is ranked #23,642 (96th percentile) for total stargazers, and #61,312 for total watchers. Github reports the primary language for this repository as Python, for repositories using this language it is ranked #3,697/126,749.

NVIDIA/TransformerEngine is also tagged with popular topics, for these it's ranked: python (#1,513/23161), deep-learning (#705/8753), machine-learning (#732/8329), pytorch (#430/6182), gpu (#78/952), cuda (#46/682)

Other Information

NVIDIA/TransformerEngine has 64 open pull requests on Github, 899 pull requests have been merged over the lifetime of the repository.

Github issues are enabled, there are 179 open issues and 228 closed issues.

There have been 16 releases, the latest one was published on 2025-02-13 (22 days ago) with the name v2.0.

Homepage URL: https://docs.nvidia.com/deeplearning/transformer-engine/user-guide/index.html

All Topics

fp8 gpu jax cuda python pytorch deep-learning machine-learning

Star History

Github stargazers over time

Watcher History

Github watchers over time, collection started in '23

Recent Commit History

914 commits on the default branch (main) since jan '22

Yearly Commits

Commits to the default branch (main) per year

Issue History

Languages

The primary language is Python but there's also others...

updated: 2025-03-07 @ 08:57pm, id: 539057023 / R_kgDOICFbfw