Starting from September 13, 2026, we will be removing the "model" type repositories of kernels (e.g., kernels-community/flash-attn3). Make sure you're using a latest version of kernels. If you face any disruption, please report them here: https://github.com/huggingface/kernels/issues/new.
KERNEL_NAME
A custom kernel for PyTorch.
Installation
pip install __REPO_ID__
Usage
import torch
from __KERNEL_NAME_NORMALIZED__ import __KERNEL_NAME_NORMALIZED__
# Create input tensor
x = torch.randn(1024, 1024, device="cuda")
# Run kernel
result = __KERNEL_NAME_NORMALIZED__(x)
Development
Building
nix develop
nix run .#build-and-copy
Testing
nix develop .#test
pytest tests/
Test as a kernels user
uv run example.py
License
Apache 2.0