feat: add mtp ut and fix some bugs (#2453)

### What this PR does / why we need it?
Fix mtp mode ut

### Does this PR introduce _any_ user-facing change?
Nothing

### How was this patch tested?
This can be tested in the same way as a unit test.


- vLLM version: v0.10.0
- vLLM main:
53415653ff

Signed-off-by: 赵江江 <zhaojiangjiang1@h-partners.com>
Co-authored-by: 赵江江 <zhaojiangjiang1@h-partners.com>

vllm_ascend/ops/fused_moe.py

@@ -1178,7 +1178,7 @@ class AscendFusedMoE(FusedMoE):
         if self.scoring_func != "softmax" and not self.use_grouped_topk:
             raise ValueError("Only softmax scoring function is supported for "
                              "non-grouped topk.")
-        self.moe = FusedMoEConfig.make(
+        moe = FusedMoEConfig.make(
             num_experts=self.global_num_experts,
             experts_per_token=top_k,
             hidden_dim=hidden_size,
@@ -1188,8 +1188,10 @@ class AscendFusedMoE(FusedMoE):
             in_dtype=params_dtype,
             quant_config=quant_config)
 
+        self.moe_config = moe
+
         if quant_config is None:
-            self.quant_method = AscendUnquantizedFusedMoEMethod(self.moe)
+            self.quant_method = AscendUnquantizedFusedMoEMethod(moe)
         else:
             self.quant_method = quant_config.get_quant_method(self, prefix)