Fix some ci issue and refactor modelrunner (#2445)

### What this PR does / why we need it?
Fix some ci issue and refactor modelrunner

### Does this PR introduce _any_ user-facing change?
N/A

### How was this patch tested?
CI passed with existing test.

- vLLM version: v0.10.0
- vLLM main:
4d9c61993a

---------

Signed-off-by: wangli <wangli858794774@gmail.com>
Signed-off-by: MengqingCao <cmq0113@163.com>
Signed-off-by: weiguihua2 <weiguihua2@huawei.com>
Co-authored-by: wangli <wangli858794774@gmail.com>
Co-authored-by: weiguihua2 <weiguihua2@huawei.com>

vllm_ascend/torchair/utils.py

@@ -2,6 +2,7 @@ import fcntl
 import os
 import shutil
 from contextlib import contextmanager, nullcontext
+from dataclasses import dataclass
 
 import torch
 
@@ -20,6 +21,32 @@ TORCHAIR_CACHE_DIR = os.getenv(
     'TORCHAIR_CACHE_HOME', os.path.join(os.getcwd(), TORCHAIR_CACHE_PATH_NAME))
 
 
+@dataclass
+class TorchairCommonAttentionMetadata:
+    """
+    Per-batch attention metadata, shared across layers and backends.
+    AttentionMetadataBuilder instances use it to construct per-layer metadata.
+    
+    For many of the tensors we keep both GPU and CPU versions.
+    """
+
+    num_reqs: int
+    """Number of requests"""
+
+    num_actual_tokens: int
+    """Total number of tokens in batch"""
+
+    decode_token_per_req: int
+
+    actual_seq_lengths_q: list[int]
+
+    attn_mask: torch.Tensor = None
+
+    spec_attn_mask: torch.Tensor = None
+
+    graph_pad_size: int = -1
+
+
 @contextmanager
 def _file_lock(file_descriptor, lock_type):
     fcntl.flock(file_descriptor, lock_type)