[Feature] Support fine-grained shared expert overlap (#5482)

Fine-grained control over shared expert overlap to prevent resource
contention.

- vLLM version: v0.13.0
- vLLM main:
5326c89803

---------

Signed-off-by: Jade Zheng <zheng.shoujian@outlook.com>

vllm_ascend/quantization/w8a8_dynamic.py

@@ -190,9 +190,6 @@ class AscendW8A8DynamicFusedMoEMethod:
         enable_force_load_balance: bool = False,
         log2phy: torch.Tensor = None,
         global_redundant_expert_num: int = 0,
-        shared_experts: Optional[Any] = None,
-        quantized_x_for_share: Optional[Any] = None,
-        dynamic_scale_for_share: Optional[Any] = None,
         pertoken_scale: Optional[Any] = None,
         **kwargs,
     ) -> torch.Tensor:
@@ -280,9 +277,6 @@ class AscendW8A8DynamicFusedMoEMethod:
             use_int8_w8a8=True,
             expert_map=expert_map,
             log2phy=log2phy,
-            shared_experts=shared_experts,
-            quantized_x_for_share=quantized_x_for_share,
-            dynamic_scale_for_share=dynamic_scale_for_share,
             dynamic_eplb=self.dynamic_eplb,
             mc2_mask=kwargs.get("mc2_mask", None))
         if zero_expert_num > 0 and zero_expert_type is not None: