[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
@@ -419,6 +419,7 @@ class QuarkW8A8Fp8MoEMethod(QuarkMoEMethod):
x: torch.Tensor,
topk_weights: torch.Tensor,
topk_ids: torch.Tensor,
shared_experts_input: torch.Tensor | None,
) -> torch.Tensor | tuple[torch.Tensor, torch.Tensor]:
if self.rocm_aiter_moe_enabled:
from vllm.model_executor.layers.fused_moe.rocm_aiter_fused_moe import (
@@ -607,6 +608,7 @@ class QuarkW4A8Fp8MoEMethod(QuarkMoEMethod):
x: torch.Tensor,
topk_weights: torch.Tensor,
topk_ids: torch.Tensor,
shared_experts_input: torch.Tensor | None,
) -> torch.Tensor | tuple[torch.Tensor, torch.Tensor]:
from vllm.model_executor.layers.fused_moe.rocm_aiter_fused_moe import (
rocm_aiter_fused_experts,
@@ -977,6 +979,7 @@ class QuarkOCP_MX_MoEMethod(QuarkMoEMethod):
x: torch.Tensor,
topk_weights: torch.Tensor,
topk_ids: torch.Tensor,
shared_experts_input: torch.Tensor | None,
) -> torch.Tensor | tuple[torch.Tensor, torch.Tensor]:
if not self.emulate:
if (