[v0.9.1][BugFix] Fix DBO bug after attn_metadata_refactor (#1445)

### What this PR does / why we need it?
fix inference bug after `set_forward_context` refactor

### Does this PR introduce _any_ user-facing change?
None

### How was this patch tested?
test case in `tests/multicard/test_torchair_graph_mode.py`

---------

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

Author: shikang-hangzhou

vllm_ascend/models/deepseek_dbo.py

@@ -34,7 +34,8 @@
 from torch import nn
 from transformers import PretrainedConfig
 from vllm.attention import Attention, AttentionMetadata
-from vllm.config import CacheConfig, ModelConfig, VllmConfig
+from vllm.config import (CacheConfig, ModelConfig, VllmConfig,
+                         get_current_vllm_config)
 from vllm.distributed import (get_pp_group,
                               get_tensor_model_parallel_world_size,
                               get_tp_group, tensor_model_parallel_all_reduce)
@@ -196,7 +197,10 @@ def __init__(
 
         self.tp_group = get_tp_group().device_group
         self.tp_rank = get_tp_group().rank_in_group
-
+        self.kv_consumer = None
+        transfer_config = get_current_vllm_config().kv_transfer_config
+        if transfer_config is not None:
+            self.kv_consumer = transfer_config.kv_role = "kv_consumer"
         self.params_dtype = torch.get_default_dtype()
 
         ascend_config = get_ascend_config()