build.toml

[general]
name = "masked-mha-runtime"
license = "Apache-2.0"
version = 1
backends = ["cuda"]

[general.cuda]
minver = "13"

[general.hub]
repo-id = "flashrt/masked-mha-runtime"

[torch]
include = ["csrc"]
src = ["torch-ext/torch_binding.cpp"]

[kernel.masked_mha_runtime]
backend = "cuda"
depends = ["torch"]
include = ["csrc"]
cuda-minver = "13"
cuda-capabilities = ["11.0a"]
cuda-flags = ["-O3", "--use_fast_math"]
src = ["csrc/attention_mha_masked.cu", "csrc/attention_mha_masked.cuh"]