[general]
name = "int8-ordered-attention"
license = "Apache-2.0"
backends = [
  "cuda",
]
version = 1
edition = 5

[general.hub]
repo-id = "wavecut/int8-ordered-attention"

[torch]
src = [
  "torch-ext/torch_binding.cpp",
  "torch-ext/torch_binding.h",
]

[kernel.int8_ordered_attention]
backend = "cuda"
depends = ["torch"]
src = ["int8_ordered_attention_cuda/attention.cu"]