This is the repository card of serizard1005/sage-attention that has been pushed on the Hub. It was built to be used with the kernels library. This card was automatically generated.

How to use

# make sure `kernels` is installed: `pip install -U kernels`
from kernels import get_kernel

kernel_module = get_kernel("serizard1005/sage-attention")
per_block_int8 = kernel_module.per_block_int8

per_block_int8(...)

Available functions

per_block_int8
per_warp_int8
sub_mean
per_channel_fp8
sageattn
sageattn3_blackwell

Benchmarks

No benchmark available yet.

Downloads last month: 21

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support