kernels-community / layer_norm

huggingface.co
Total runs: 12
24-hour runs: 0
7-day runs: 11
30-day runs: -35
Model's Last Updated: 2026年3月5日

Introduction of layer_norm

Model Details of layer_norm

This CUDA extension implements fused dropout + residual + LayerNorm, building on Apex's FastLayerNorm . Major changes:

  • Add dropout and residual.
  • Make it work for both pre-norm and post-norm architecture.
  • Support more hidden dimensions (all dimensions divisible by 8, up to 8192).
  • Implement RMSNorm as an option.
  • Support layer norm with parallel residual (e.g., GPT-J, GPT-NeoX, PaLM).

If you want to use it for dimensions larger than 8k, please file an issue.

This extension has only been tested on A100s.

cd csrc/layer_norm && pip install .

As of 2024-01-05, this extension is no longer used in the FlashAttention repo. We've instead switched to a Triton-based implementation .

Runs of kernels-community layer_norm on huggingface.co

12
Total runs
0
24-hour runs
-1
3-day runs
11
7-day runs
-35
30-day runs

More Information About layer_norm huggingface.co Model

layer_norm huggingface.co

layer_norm huggingface.co is an AI model on huggingface.co that provides layer_norm's model effect (), which can be used instantly with this kernels-community layer_norm model. huggingface.co supports a free trial of the layer_norm model, and also provides paid use of the layer_norm. Support call layer_norm model through api, including Node.js, Python, http.

kernels-community layer_norm online free

layer_norm huggingface.co is an online trial and call api platform, which integrates layer_norm's modeling effects, including api services, and provides a free online trial of layer_norm, you can try layer_norm online for free by clicking the link below.

kernels-community layer_norm online free url in huggingface.co:

https://huggingface.co/kernels-community/layer_norm

layer_norm install

layer_norm is an open source model from GitHub that offers a free installation service, and any user can find layer_norm on GitHub to install. At the same time, huggingface.co provides the effect of layer_norm install, users can directly use layer_norm installed effect in huggingface.co for debugging and trial. It also supports api for free installation.

layer_norm install url in huggingface.co:

https://huggingface.co/kernels-community/layer_norm

Url of layer_norm

Provider of layer_norm huggingface.co

kernels-community
ORGANIZATIONS

Other API from kernels-community

huggingface.co

Total runs: 44.0K
Run Growth: -15.0K
Growth Rate: -34.06%
Updated:2026年6月4日
huggingface.co

Total runs: 2.9K
Run Growth: 2.7K
Growth Rate: 93.75%
Updated:2026年5月5日