[3/N][Refactor][Quantization]remove packed_modules_mapping from models (#3021)

### What this PR does / why we need it?

Some custom models in vllm-ascend define packed_modules_mapping, which
prevent keeping same model class with vllm community. So move these
custom packed_modules_mapping to quant utils.py. After this pr, some
custom models can be removed.

### Does this PR introduce _any_ user-facing change?

tested by CI

### How was this patch tested?

tested by CI

- vLLM version: v0.10.2
- vLLM main:
5089fd749c

Signed-off-by: 22dimensions <waitingwind@foxmail.com>

tests/ut/models/test_qwen3_moe.py

@@ -15,41 +15,11 @@
 import math
 import unittest
 
-import pytest
 import torch
-from vllm.model_executor.models.qwen3_moe import Qwen3MoeForCausalLM
 
-from vllm_ascend.models.qwen3_moe import CustomQwen3MoeForCausalLM
 from vllm_ascend.torchair.models.qwen3_moe import CustomQwen3MoeAttention
 
 
-class TestCustomQwen3MoeForCausalLM:
-
-    def test_class_inheritance(self):
-        assert issubclass(CustomQwen3MoeForCausalLM, Qwen3MoeForCausalLM)
-
-    @pytest.mark.parametrize("key, expected", [
-        ("qkv_proj", ["q_proj", "k_proj", "v_proj"]),
-        ("gate_up_proj", ["gate_proj", "up_proj"]),
-        ("experts",
-         ["experts.0.gate_proj", "experts.0.up_proj", "experts.0.down_proj"]),
-    ])
-    def test_packed_modules_mapping(self, key, expected):
-        assert CustomQwen3MoeForCausalLM.packed_modules_mapping[
-            key] == expected
-
-    def test_packed_modules_mapping_structure(self):
-        expected_mapping = {
-            "qkv_proj": ["q_proj", "k_proj", "v_proj"],
-            "gate_up_proj": ["gate_proj", "up_proj"],
-            "experts": [
-                "experts.0.gate_proj", "experts.0.up_proj",
-                "experts.0.down_proj"
-            ]
-        }
-        assert CustomQwen3MoeForCausalLM.packed_modules_mapping == expected_mapping
-
-
 class DummyRMSNorm:
 
     def __init__(self, dim: int, eps: float = 1e-6):