NVIDIA / TransformerEngine
A library for accelerating Transformer models on NVIDIA GPUs, including using 8-bit and 4-bit floating point (FP8 and FP4) precision on Hopper, Ada and Blackwell GPUs, to provide better performance with lower memory utilization in both training and inference.
Why this job: matched by what the repo talks about: inference, gpu, model, performance
What it does
Copyright (c) 2022-2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
Get started
No single install command. Read the steps in the README (opens in new tab)
- Maintenance
- Active
- Last commit today
- Stars
- 3,603
- +6 this week
- Licence
- Apache-2.0
- Allows business use
- Downloads for
- Linux
- From the latest release
- Latest version
- v2.21.0
- Released 2 days ago
- Built with
- Python
- 61% of the code
Is it looked after?
- Contributors
- 44
- Last 90 days
- Issues
- 81% closed
- 156 still open
- Pull requests
- 2,312 merged
- 230 waiting
- Releases
- 40
- 15 in the last year
- Contributing guide (opens in new tab)
- Security policy (opens in new tab)
- No code of conduct
Similar repos
| Repo | Stars | Last commit | Licence | Downloads for |
|---|---|---|---|---|
| flashinfer-ai / flashinfer | 6,900 | today | Apache-2.0 | Linux |
| AI-Hypercomputer / JetStream | 473 | 10 months ago | Apache-2.0 | No downloads |
| jax-ml / jax-llm-examples | 287 | 2 months ago | Apache-2.0 | No downloads |
More: download files, languages, topics and project record
Languages
- Python 61%
- Cuda 24%
- C++ 13%
- C 1%
Topics
- cuda
- deep-learning
- gpu
- machine-learning
- python
- pytorch
- fp8
- jax
- fp4
Project record
- JobLocal LLMs
- StatusListed
- Commits2,341
- Forks894
- Watching35
- Stars in 28 days+28
- CreatedSep 20, 2022
- First seen by usNov 7, 2026
- CheckedNov 12, 2026