[MoE Refactor] Introduce MoERunner abstraction and move execution logic from FusedMoE to DefaultMoERunner (#32344)

Signed-off-by: Bill Nell <bnell@redhat.com>
This commit is contained in:
bnellnm
2026-02-10 19:51:07 -05:00
committed by GitHub
parent dc6de33c3d
commit d1481ba783
25 changed files with 913 additions and 753 deletions
@@ -958,6 +958,7 @@ class ModelOptFp8MoEMethod(FusedMoEMethodBase):
x: torch.Tensor,
topk_weights: torch.Tensor,
topk_ids: torch.Tensor,
shared_experts_input: torch.Tensor | None,
) -> torch.Tensor | tuple[torch.Tensor, torch.Tensor]:
assert not self.is_monolithic
@@ -980,7 +981,7 @@ class ModelOptFp8MoEMethod(FusedMoEMethodBase):
global_num_experts=layer.global_num_experts,
expert_map=layer.expert_map,
apply_router_weight_on_input=layer.apply_router_weight_on_input,
shared_experts_input=layer._get_shared_experts_input(x),
shared_experts_input=shared_experts_input,
)
@@ -1524,6 +1525,7 @@ class ModelOptNvFp4FusedMoE(FusedMoEMethodBase):
x: torch.Tensor,
topk_weights: torch.Tensor,
topk_ids: torch.Tensor,
shared_experts_input: torch.Tensor | None,
) -> torch.Tensor | tuple[torch.Tensor, torch.Tensor]:
assert not self.is_monolithic
@@ -1551,7 +1553,7 @@ class ModelOptNvFp4FusedMoE(FusedMoEMethodBase):
global_num_experts=layer.global_num_experts,
expert_map=layer.expert_map,
apply_router_weight_on_input=layer.apply_router_weight_on_input,
shared_experts_input=layer._get_shared_experts_input(x),
shared_experts_input=shared_experts_input,
)