fix codecheck

Signed-off-by: David9857 <985700846@qq.com>

Author: David9857

vllm_ascend/models/deepseek_v2.py

@@ -30,7 +30,7 @@
 import torch
 import torch.distributed as dist
 import torch_npu
-import torchair as tng
+import torchair as tng  # type: ignore
 import vllm.envs as envs
 from torch import nn
 from transformers import PretrainedConfig
@@ -257,7 +257,8 @@ def forward(self, hidden_states: torch.Tensor) -> torch.Tensor:
         if self.n_shared_experts is not None and cv_parallel:
             with tng.scope.npu_stream_switch('cv'):
                 tng.scope.npu_wait_tensor(shared_hidden_states, router_logits)
-                x, dynamic_scale = torch_npu.npu_dynamic_quant(shared_hidden_states)
+                x, dynamic_scale = torch_npu.npu_dynamic_quant(
+                    shared_hidden_states)
                 gate_up = torch_npu.npu_quant_matmul(
                     x,
                     self.shared_experts.gate_up_proj.weight,
@@ -274,8 +275,7 @@ def forward(self, hidden_states: torch.Tensor) -> torch.Tensor:
                 enable_force_load_balance=enable_force_load_balance,
                 shared_experts=self.shared_experts,
                 shared_gate_up=gate_up,
-                shared_dynamic_scale=dynamic_scale
-            )
+                shared_dynamic_scale=dynamic_scale)
             router_hidden_states = router_hidden_states * self.routed_scaling_factor
         else:
             router_hidden_states = self.experts(

vllm_ascend/quantization/w8a8_dynamic.py

@@ -20,7 +20,7 @@
 import torch
 import torch.distributed as dist
 import torch_npu
-import torchair as tng
+import torchair as tng  # type: ignore
 from vllm.distributed import GroupCoordinator, tensor_model_parallel_all_reduce
 
 import vllm_ascend.envs as envs_ascend
@@ -133,19 +133,17 @@ def apply_mlp(hidden_states_wrapper: List[torch.Tensor],
     return hidden_states
 
 
-def fused_experts_with_mc2(
-    hidden_states: torch.Tensor,
-    w1: torch.Tensor,
-    w2: torch.Tensor,
-    w1_scale: torch.Tensor,
-    w2_scale: torch.Tensor,
-    topk_weights: torch.Tensor,
-    topk_ids: torch.Tensor,
-    top_k: int,
-    expert_map: torch.Tensor = None,
-    moe_all_to_all_group_name: str = "",
-    **kwargs
-) -> torch.Tensor:
+def fused_experts_with_mc2(hidden_states: torch.Tensor,
+                           w1: torch.Tensor,
+                           w2: torch.Tensor,
+                           w1_scale: torch.Tensor,
+                           w2_scale: torch.Tensor,
+                           topk_weights: torch.Tensor,
+                           topk_ids: torch.Tensor,
+                           top_k: int,
+                           expert_map: torch.Tensor = None,
+                           moe_all_to_all_group_name: str = "",
+                           **kwargs) -> torch.Tensor:
     global_bs = 0
     moe_expert_num = len(expert_map)
     # hidden_states = hidden_states.bfloat16()