[EP] Query NIXL EP top-k index dtype (#45298)

Signed-off-by: Itay Alroy <ialroy@nvidia.com>
This commit is contained in:
Itay Alroy
2026-06-15 22:50:18 +00:00
committed by GitHub
parent ab8b0fe338
commit 16e91176cf
@@ -28,6 +28,8 @@ logger = init_logger(__name__)
# NIXL EP kernels quantize dispatch inputs in 128 element chunks.
NIXL_EP_QUANT_BLOCK_SIZE = 128
NIXL_EP_QUANT_BLOCK_SHAPE = [NIXL_EP_QUANT_BLOCK_SIZE, NIXL_EP_QUANT_BLOCK_SIZE]
NIXL_EP_TOPK_INDICES_DTYPE = getattr(nixl_ep, "topk_idx_t", torch.int64)
assert isinstance(NIXL_EP_TOPK_INDICES_DTYPE, torch.dtype)
def dequant_fp8(
@@ -151,7 +153,7 @@ class NixlEPPrepareAndFinalize(mk.FusedMoEPrepareAndFinalizeModular):
all2all_manager.commit_staged_state()
def topk_indices_dtype(self) -> torch.dtype | None:
return torch.int64
return NIXL_EP_TOPK_INDICES_DTYPE
def _map_global_to_physical_ids(self, topk_ids: torch.Tensor) -> torch.Tensor:
if self.global_to_physical is None: