diff --git a/mindarmour/diff_privacy/train/model.py b/mindarmour/diff_privacy/train/model.py
index dcd02b17165bfa64db09ddbc3aba4c3fbd7ec837..3243de53eaa17838c1089515ac9839b3231484bd 100644
--- a/mindarmour/diff_privacy/train/model.py
+++ b/mindarmour/diff_privacy/train/model.py
@@ -143,7 +143,13 @@ class DPModel(Model):
                 LOGGER.error(TAG, msg)
                 raise ValueError(msg)
         self._noise_mech = noise_mech
-        if clip_mech is not None:
+        if noise_mech is not None:
+            if 'Ada' in noise_mech.__class__.__name__ and clip_mech is not None:
+                msg = 'When noise_mech is Adaptive, clip_mech must be None.'
+                LOGGER.error(TAG, msg)
+                raise ValueError(msg)
+
+        if clip_mech is None or isinstance(clip_mech, Cell):
             self._clip_mech = clip_mech
         super(DPModel, self).__init__(**kwargs)
 
diff --git a/tests/ut/python/diff_privacy/test_model_train.py b/tests/ut/python/diff_privacy/test_model_train.py
index 892913c45182a128bd59cccd3cf6aa2e71a81df0..6a4b716de3e58a9b7fea3731b30c6f7ff57c0e19 100644
--- a/tests/ut/python/diff_privacy/test_model_train.py
+++ b/tests/ut/python/diff_privacy/test_model_train.py
@@ -134,11 +134,7 @@ def test_dp_model_with_graph_mode_ada_gaussian():
     noise_mech = NoiseMechanismsFactory().create('AdaGaussian',
                                                  norm_bound=norm_clip,
                                                  initial_noise_multiplier=initial_noise_multiplier)
-    clip_mech = ClipMechanismsFactory().create('Gaussian',
-                                               decay_policy='Linear',
-                                               learning_rate=0.01,
-                                               target_unclipped_quantile=0.9,
-                                               fraction_stddev=0.01)
+    clip_mech = None
     net_opt = nn.Momentum(network.trainable_params(), learning_rate=0.1,
                           momentum=0.9)
     model = DPModel(micro_batches=2,