diff --git a/orttraining/orttraining/python/training/_ortmodule_graph_execution_manager.py b/orttraining/orttraining/python/training/_ortmodule_graph_execution_manager.py
index 6f10e5bcd7..6f99d68044 100644
--- a/orttraining/orttraining/python/training/_ortmodule_graph_execution_manager.py
+++ b/orttraining/orttraining/python/training/_ortmodule_graph_execution_manager.py
@@ -9,6 +9,8 @@ from . import _ortmodule_logger as _logger
 from onnxruntime.capi.onnxruntime_inference_collection import OrtValue
 from onnxruntime.capi import _pybind_state as C
 
+from onnxruntime.tools.symbolic_shape_infer import SymbolicShapeInference
+
 from abc import ABC, abstractmethod
 import io
 import inspect
@@ -90,6 +92,8 @@ class GraphExecutionManager(ABC):
         # default execution order is priority-based for both dynamic/static shape input for now
         # if we observe benefit of static shape, we can expose this flag to user
         self._use_static_shape = False
+        # flag to enable symbolic shape inference for dynamic shape inputs to improve performance
+        self._run_symbolic_shape_infer = False
         self._input_names_require_grad = None
         self._module_output_schema = None
 
@@ -184,6 +188,9 @@ class GraphExecutionManager(ABC):
         self._set_device_from_module()
         self._onnx_model = self._get_exported_model(*inputs, **kwargs)
 
+        if self._run_symbolic_shape_infer:
+            self._onnx_model = SymbolicShapeInference.infer_shapes(self._onnx_model, auto_merge=True, guess_output_rank=True)
+
         return True
 
     def _get_exported_model(self, *inputs, **kwargs):