clean code

Signed-off-by: shikang-hangzhou <459956190@qq.com>

Author: shikang-hangzhou

vllm_ascend/multistream/ms_split.py

@@ -72,11 +72,11 @@ def model_input_split_v1_mla_attn(
             attn_metadata.query_lens):
         return [attn_metadata]
 
-    query_start_loc_cpu = np.zeros(shape=(len(attn_metadata.query_lens) + 1, ),
+    query_start_loc_cpu: Any = np.zeros(shape=(len(attn_metadata.query_lens) + 1, ),
                                    dtype=int)
     np.cumsum(attn_metadata.query_lens, out=query_start_loc_cpu[1:])
     if attn_metadata.num_prefills > 0:
-        prefill_query_start_loc = np.zeros(
+        prefill_query_start_loc: Any = np.zeros(
             shape=(len(attn_metadata.prefill.query_lens) + 1, ), dtype=int)
         np.cumsum(attn_metadata.prefill.query_lens,
                   out=prefill_query_start_loc[1:])

vllm_ascend/worker/model_runner_v1.py

@@ -26,7 +26,7 @@
 import weakref
 from contextlib import contextmanager, nullcontext
 from dataclasses import dataclass
-from typing import TYPE_CHECKING, Dict, List, Optional, Union
+from typing import TYPE_CHECKING, Any, Dict, List, Optional, Union
 
 import numpy as np
 import numpy.typing as npt
@@ -1160,7 +1160,7 @@ def _calc_spec_decode_metadata(
 
         # Compute the logits indices.
         # [4, 1, 3, 1, 2]
-        num_sampled_tokens = num_draft_tokens + 1
+        num_sampled_tokens: Any = num_draft_tokens + 1
         # Step 1. [4, 5, 8, 9, 11]
         cu_num_sampled_tokens = np.cumsum(num_sampled_tokens, dtype=np.int32)
         total_num_sampled_tokens = cu_num_sampled_tokens[-1]