UT fix

Co-authored-by: chenxiao <[email protected]>
Signed-off-by: zzhx1 <[email protected]>

Author: zzhx1

tests/ut/attention/test_mla_v1.py

@@ -6,6 +6,7 @@
 from vllm.model_executor.layers.linear import LinearBase
 
 from tests.ut.base import TestBase
+from vllm_ascend.ascend_config import init_ascend_config
 from vllm_ascend.attention.attention_v1 import AscendAttentionState
 from vllm_ascend.attention.mla_v1 import (AscendMLABackend,
                                           AscendMLADecodeMetadata,
@@ -723,6 +724,13 @@ def setUp(self, ascend_config, get_current_vllm_config, mock_get_tp_size,
         model_config.dtype = torch.float16
         vllm_config.model_config = model_config
         get_current_vllm_config.return_value = vllm_config
+        vllm_config.additional_config = {
+            "torchair_graph_config": {
+                "enabled": False,
+            },
+            "refresh": True
+        }
+        init_ascend_config(vllm_config)
 
         num_heads = 256
         head_size = 1024