Fix Llama4 FlashInfer FP4 MoE issues (#22511)

Signed-off-by: Po-Han Huang <pohanh@nvidia.com>

vllm/model_executor/layers/fused_moe/flashinfer_cutlass_moe.py

@@ -170,8 +170,6 @@ class FlashInferExperts(mk.FusedMoEPermuteExpertsUnpermute):
             "w1_scale and w2_scale must not "
             "be None for FlashInferExperts")
 
-        assert not apply_router_weight_on_input
-
         quant_scales = [
             a1_gscale,
             w1_scale.view(torch.int32),

vllm/model_executor/layers/fused_moe/flashinfer_cutlass_prepare_finalize.py

@@ -60,7 +60,12 @@ class FlashInferCutlassMoEPrepareAndFinalize(mk.FusedMoEPrepareAndFinalize):
     ) -> tuple[torch.Tensor, Optional[torch.Tensor], Optional[torch.Tensor],
                Optional[torch.Tensor], Optional[torch.Tensor]]:
 
-        assert not apply_router_weight_on_input
+        if apply_router_weight_on_input:
+            topk = topk_ids.size(1)
+            # TODO: this only works for topK=1, will need to update for topK>1
+            assert topk == 1, \
+                "apply_router_weight_on_input is only implemented for topk=1"
+            a1.mul_(topk_weights.to(a1.dtype))
 
         (a1_gscale, use_dp, local_tokens) = extract_required_args(
             extra_prepare_args, ['a1_gscale', 'use_dp', 'local_tokens'])

vllm/model_executor/layers/quantization/modelopt.py

@@ -1299,8 +1299,9 @@ class ModelOptNvFp4FusedMoE(FusedMoEMethodBase):
                 output2_scale_scalar=layer.g2_alphas.data,
                 num_experts=global_num_experts,
                 top_k=top_k,
-                n_group=num_expert_group,
-                topk_group=topk_group,
+                n_group=num_expert_group
+                if num_expert_group is not None else 0,
+                topk_group=topk_group if topk_group is not None else 0,
                 intermediate_size=layer.intermediate_size_per_partition,
                 local_expert_offset=layer.ep_rank * layer.local_num_experts,
                 local_num_experts=layer.local_num_experts,