[Hardware][NVIDIA] FP4 MoE kernel optimization (#19110)
Signed-off-by: Chiyue Wei <chiyuew@nvidia.com> Co-authored-by: Chiyue Wei <chiyuew@nvidia.com>
This commit is contained in:
@@ -81,6 +81,12 @@ TORCH_LIBRARY_EXPAND(TORCH_EXTENSION_NAME, m) {
|
||||
m.def("moe_permute_unpermute_supported() -> bool");
|
||||
m.impl("moe_permute_unpermute_supported", &moe_permute_unpermute_supported);
|
||||
|
||||
// Row shuffle for MoE
|
||||
m.def(
|
||||
"shuffle_rows(Tensor input_tensor, Tensor dst2src_map, Tensor! "
|
||||
"output_tensor) -> ()");
|
||||
m.impl("shuffle_rows", torch::kCUDA, &shuffle_rows);
|
||||
|
||||
#endif
|
||||
}
|
||||
|
||||
|
||||
Reference in New Issue
Block a user