diff --git a/examples/privacy/diff_privacy/lenet5_dp_ada_gaussian.py b/examples/privacy/diff_privacy/lenet5_dp_ada_gaussian.py
index 918d824..b8aa663 100644
--- a/examples/privacy/diff_privacy/lenet5_dp_ada_gaussian.py
+++ b/examples/privacy/diff_privacy/lenet5_dp_ada_gaussian.py
@@ -122,7 +122,8 @@ if __name__ == "__main__":
                                                num_samples=60000,
                                                batch_size=cfg.batch_size,
                                                initial_noise_multiplier=cfg.initial_noise_multiplier,
-                                               per_print_times=234)
+                                               per_print_times=234,
+                                               target_delta=1e-5)
     # Create the DP model for training.
     model = DPModel(micro_batches=cfg.micro_batches,
                     norm_bound=cfg.norm_bound,
diff --git a/examples/privacy/diff_privacy/lenet5_dp_ada_sgd_graph.py b/examples/privacy/diff_privacy/lenet5_dp_ada_sgd_graph.py
index 0c36419..0baf539 100644
--- a/examples/privacy/diff_privacy/lenet5_dp_ada_sgd_graph.py
+++ b/examples/privacy/diff_privacy/lenet5_dp_ada_sgd_graph.py
@@ -122,7 +122,8 @@ if __name__ == "__main__":
                                                num_samples=60000,
                                                batch_size=cfg.batch_size,
                                                initial_noise_multiplier=cfg.initial_noise_multiplier,
-                                               per_print_times=234)
+                                               per_print_times=234,
+                                               target_delta=1e-5)
     # Create the DP model for training.
     model = DPModel(micro_batches=cfg.micro_batches,
                     norm_bound=cfg.norm_bound,
diff --git a/examples/privacy/diff_privacy/lenet5_dp_optimizer.py b/examples/privacy/diff_privacy/lenet5_dp_optimizer.py
index 05f7205..db38965 100644
--- a/examples/privacy/diff_privacy/lenet5_dp_optimizer.py
+++ b/examples/privacy/diff_privacy/lenet5_dp_optimizer.py
@@ -129,7 +129,8 @@ if __name__ == "__main__":
                                                num_samples=60000,
                                                batch_size=cfg.batch_size,
                                                initial_noise_multiplier=cfg.initial_noise_multiplier*cfg.norm_bound,
-                                               per_print_times=10)
+                                               per_print_times=10,
+                                               target_delta=1e-5)
     # Create the DP model for training.
     model = DPModel(micro_batches=cfg.micro_batches,
                     norm_bound=cfg.norm_bound,