Merge branch 'v0.9.2-dev-fth' into 'v0.9.2-dev'

fp8增加fused_moe_gate参数

See merge request dcutoolkit/deeplearing/vllm!391

vllm/model_executor/layers/quantization/fp8.py

@@ -858,6 +858,7 @@ class Fp8MoEMethod(FusedMoEMethodBase):
         e_score_correction_bias: Optional[torch.Tensor] = None,
         apply_router_weight_on_input: bool = False,
         activation: str = "silu",
+        use_fused_gate: Optional[bool] = False,
         enable_eplb: bool = False,
         expert_load_view: Optional[torch.Tensor] = None,
         logical_to_physical_map: Optional[torch.Tensor] = None,
@@ -886,6 +887,7 @@ class Fp8MoEMethod(FusedMoEMethodBase):
             expert_load_view=expert_load_view,
             logical_to_physical_map=logical_to_physical_map,
             logical_replica_count=logical_replica_count,
+            use_fused_gate=use_fused_gate,
         )
 
         if self.rocm_aiter_moe_enabled: