[Qwen3VL] Add clear_grpah_opt_backend method to Qwen3VLForConditionalGeneration (#7086)

Add clear_grpah_opt_backend method that delegates to the underlying model
to clear cuda graph optimization backend.

Co-authored-by: CSWYF3634076 <wangyafeng@baidu.com>

fastdeploy/model_executor/models/qwen3_vl/qwen3_vl.py

@@ -382,6 +382,10 @@ class Qwen3VLForConditionalGeneration(ModelForCasualLM):
 
         return hidden_states
 
+    def clear_grpah_opt_backend(self):
+        """Clear graph optimization backend, the captured cuda graph will be cleaned"""
+        self.model.clear_grpah_opt_backend(fd_config=self.fd_config)
+
 
 class Qwen3VLPretrainedModel(PretrainedModel):
     """Utilities for tensor-parallel weight splitting."""