add norm 2.0 api, test=develop (#26465)

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

* add norm 2.0 api, test=develop

python/paddle/fluid/tests/unittests/test_batch_norm_op_v2.py

+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import os
+import unittest
+import numpy as np
+import paddle.fluid.core as core
+from paddle.fluid.op import Operator
+import paddle.fluid as fluid
+from op_test import OpTest, _set_use_system_allocator
+from paddle.fluid.framework import grad_var_name
+import paddle.fluid as fluid
+from paddle.fluid import Program, program_guard
+import paddle
+
+
+class TestBatchNorm(unittest.TestCase):
+    def test_name(self):
+        places = [fluid.CPUPlace()]
+        if core.is_compiled_with_cuda() and core.op_support_gpu("batch_norm"):
+            places.append(fluid.CUDAPlace(0))
+        for p in places:
+            with fluid.dygraph.guard(p):
+                batch_norm1d = paddle.nn.BatchNorm1d(1, name="test")
+
+    def test_error(self):
+        places = [fluid.CPUPlace()]
+        if core.is_compiled_with_cuda() and core.op_support_gpu("batch_norm"):
+            places.append(fluid.CUDAPlace(0))
+        for p in places:
+            #paddle.disable_static()
+            x_data_4 = np.random.random(size=(2, 1, 3, 3)).astype('float32')
+            x_data_3 = np.random.random(size=(2, 1, 3)).astype('float32')
+
+            def error1d():
+                x_data_4 = np.random.random(size=(2, 1, 3, 3)).astype('float32')
+                batch_norm1d = paddle.nn.BatchNorm1d(1)
+                batch_norm1d(fluid.dygraph.to_variable(x_data_4))
+
+            def error2d():
+                x_data_3 = np.random.random(size=(2, 1, 3)).astype('float32')
+                batch_norm2d = paddle.nn.BatchNorm2d(1)
+                batch_norm2d(fluid.dygraph.to_variable(x_data_3))
+
+            def error3d():
+                x_data_4 = np.random.random(size=(2, 1, 3, 3)).astype('float32')
+                batch_norm3d = paddle.nn.BatchNorm3d(1)
+                batch_norm3d(fluid.dygraph.to_variable(x_data_4))
+
+            with fluid.dygraph.guard(p):
+                self.assertRaises(ValueError, error1d)
+                self.assertRaises(ValueError, error2d)
+                self.assertRaises(ValueError, error3d)
+
+    def test_dygraph(self):
+        places = [fluid.CPUPlace()]
+        if core.is_compiled_with_cuda() and core.op_support_gpu("batch_norm"):
+            places.append(fluid.CUDAPlace(0))
+        for p in places:
+            shape = [4, 10, 4, 4]
+
+            def compute_v1(x, is_test, trainable_statistics):
+                with fluid.dygraph.guard(p):
+                    bn = fluid.dygraph.BatchNorm(
+                        shape[1],
+                        is_test=is_test,
+                        trainable_statistics=trainable_statistics)
+                    y = bn(fluid.dygraph.to_variable(x))
+                return y.numpy()
+
+            def compute_v2(x):
+                with fluid.dygraph.guard(p):
+                    bn = paddle.nn.BatchNorm2d(shape[1])
+                    y = bn(fluid.dygraph.to_variable(x))
+                return y.numpy()
+
+            x = np.random.randn(*shape).astype("float32")
+            y1 = compute_v1(x, False, False)
+            y2 = compute_v2(x)
+            self.assertTrue(np.allclose(y1, y2))
+
+    def test_static(self):
+        places = [fluid.CPUPlace()]
+        if core.is_compiled_with_cuda() and core.op_support_gpu("batch_norm"):
+            places.append(fluid.CUDAPlace(0))
+        for p in places:
+            exe = fluid.Executor(p)
+            shape = [4, 10, 16, 16]
+
+            def compute_v1(x_np, is_test, trainable_statistics):
+                with program_guard(Program(), Program()):
+                    bn = fluid.dygraph.BatchNorm(
+                        shape[1],
+                        is_test=is_test,
+                        trainable_statistics=trainable_statistics)
+                    x = fluid.data(name='x', shape=x_np.shape, dtype=x_np.dtype)
+                    y = bn(x)
+                    exe.run(fluid.default_startup_program())
+                    r = exe.run(feed={'x': x_np}, fetch_list=[y])[0]
+                return r
+
+            def compute_v2(x_np):
+                with program_guard(Program(), Program()):
+                    bn = paddle.nn.BatchNorm2d(shape[1])
+                    x = fluid.data(name='x', shape=x_np.shape, dtype=x_np.dtype)
+                    y = bn(x)
+                    exe.run(fluid.default_startup_program())
+                    r = exe.run(feed={'x': x_np}, fetch_list=[y])[0]
+                return r
+
+            x = np.random.randn(*shape).astype("float32")
+            y1 = compute_v1(x, False, False)
+            y2 = compute_v2(x)
+            self.assertTrue(np.allclose(y1, y2))
+
+
+if __name__ == '__main__':
+    unittest.main()

python/paddle/fluid/tests/unittests/test_group_norm_op_v2.py

+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import os
+import unittest
+import numpy as np
+import paddle.fluid.core as core
+from paddle.fluid.op import Operator
+import paddle.fluid as fluid
+from op_test import OpTest, _set_use_system_allocator
+from paddle.fluid.framework import grad_var_name
+import paddle.fluid as fluid
+from paddle.fluid import Program, program_guard
+import paddle
+
+
+class TestDygraphGroupNormv2(unittest.TestCase):
+    def test_dygraph(self):
+        places = [fluid.CPUPlace()]
+        if core.is_compiled_with_cuda() and core.op_support_gpu("group_norm"):
+            places.append(fluid.CUDAPlace(0))
+        for p in places:
+            shape = [2, 6, 2, 2]
+
+            def compute_v1(x):
+                with fluid.dygraph.guard(p):
+                    gn = fluid.dygraph.GroupNorm(channels=2, groups=2)
+                    y = gn(fluid.dygraph.to_variable(x))
+                return y.numpy()
+
+            def compute_v2(x):
+                with fluid.dygraph.guard(p):
+                    gn = paddle.nn.GroupNorm(num_channels=2, num_groups=2)
+                    y = gn(fluid.dygraph.to_variable(x))
+                return y.numpy()
+
+            x = np.random.randn(*shape).astype("float32")
+            y1 = compute_v1(x)
+            y2 = compute_v2(x)
+            self.assertTrue(np.allclose(y1, y2))
+
+    def test_static(self):
+        places = [fluid.CPUPlace()]
+        if core.is_compiled_with_cuda() and core.op_support_gpu("layer_norm"):
+            places.append(fluid.CUDAPlace(0))
+        for p in places:
+            exe = fluid.Executor(p)
+            shape = [2, 6, 2, 2]
+
+            def compute_v1(x_np):
+                with program_guard(Program(), Program()):
+                    gn = fluid.dygraph.GroupNorm(channels=2, groups=2)
+                    x = fluid.data(name='x', shape=x_np.shape, dtype=x_np.dtype)
+                    y = gn(x)
+                    exe.run(fluid.default_startup_program())
+                    r = exe.run(feed={'x': x_np}, fetch_list=[y])[0]
+                return r
+
+            def compute_v2(x_np):
+                with program_guard(Program(), Program()):
+                    gn = paddle.nn.GroupNorm(num_channels=2, num_groups=2)
+                    x = fluid.data(name='x', shape=x_np.shape, dtype=x_np.dtype)
+                    y = gn(x)
+                    exe.run(fluid.default_startup_program())
+                    r = exe.run(feed={'x': x_np}, fetch_list=[y])[0]
+                return r
+
+            x = np.random.randn(*shape).astype("float32")
+            y1 = compute_v1(x)
+            y2 = compute_v2(x)
+            self.assertTrue(np.allclose(y1, y2))
+
+
+if __name__ == '__main__':
+    unittest.main()

python/paddle/fluid/tests/unittests/test_instance_norm_op_v2.py

+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import os
+import unittest
+import numpy as np
+import paddle.fluid.core as core
+from paddle.fluid.op import Operator
+import paddle.fluid as fluid
+from op_test import OpTest, _set_use_system_allocator
+from paddle.fluid.framework import grad_var_name
+import paddle.fluid as fluid
+from paddle.fluid import Program, program_guard
+import paddle
+
+
+class TestInstanceNorm(unittest.TestCase):
+    def test_error(self):
+        places = [fluid.CPUPlace()]
+        if core.is_compiled_with_cuda() and core.op_support_gpu(
+                "instance_norm"):
+            places.append(fluid.CUDAPlace(0))
+        for p in places:
+
+            def error1d():
+                x_data_4 = np.random.random(size=(2, 1, 3, 3)).astype('float32')
+                instance_norm1d = paddle.nn.InstanceNorm1d(1)
+                instance_norm1d(fluid.dygraph.to_variable(x_data_4))
+
+            def error2d():
+                x_data_3 = np.random.random(size=(2, 1, 3)).astype('float32')
+                instance_norm2d = paddle.nn.InstanceNorm2d(1)
+                instance_norm2d(fluid.dygraph.to_variable(x_data_3))
+
+            def error3d():
+                x_data_4 = np.random.random(size=(2, 1, 3, 3)).astype('float32')
+                instance_norm3d = paddle.nn.BatchNorm3d(1)
+                instance_norm3d(fluid.dygraph.to_variable(x_data_4))
+
+            with fluid.dygraph.guard(p):
+                self.assertRaises(ValueError, error1d)
+                self.assertRaises(ValueError, error2d)
+                self.assertRaises(ValueError, error3d)
+
+    def test_dygraph(self):
+        places = [fluid.CPUPlace()]
+        if core.is_compiled_with_cuda() and core.op_support_gpu(
+                "instance_norm"):
+            places.append(fluid.CUDAPlace(0))
+        for p in places:
+            shape = [4, 10, 4, 4]
+
+            def compute_v1(x):
+                with fluid.dygraph.guard(p):
+                    bn = fluid.dygraph.InstanceNorm(shape[1])
+                    y = bn(fluid.dygraph.to_variable(x))
+                return y.numpy()
+
+            def compute_v2(x):
+                with fluid.dygraph.guard(p):
+                    bn = paddle.nn.InstanceNorm2d(shape[1])
+                    y = bn(fluid.dygraph.to_variable(x))
+                return y.numpy()
+
+            x = np.random.randn(*shape).astype("float32")
+            y1 = compute_v1(x)
+            y2 = compute_v2(x)
+            self.assertTrue(np.allclose(y1, y2))
+
+    def test_static(self):
+        places = [fluid.CPUPlace()]
+        if core.is_compiled_with_cuda() and core.op_support_gpu(
+                "instance_norm"):
+            places.append(fluid.CUDAPlace(0))
+        for p in places:
+            exe = fluid.Executor(p)
+            shape = [4, 10, 16, 16]
+
+            def compute_v1(x_np):
+                with program_guard(Program(), Program()):
+                    ins = fluid.dygraph.InstanceNorm(shape[1])
+                    x = fluid.data(name='x', shape=x_np.shape, dtype=x_np.dtype)
+                    y = ins(x)
+                    exe.run(fluid.default_startup_program())
+                    r = exe.run(feed={'x': x_np}, fetch_list=[y])[0]
+                return r
+
+            def compute_v2(x_np):
+                with program_guard(Program(), Program()):
+                    ins = paddle.nn.InstanceNorm2d(shape[1])
+                    x = fluid.data(name='x', shape=x_np.shape, dtype=x_np.dtype)
+                    y = ins(x)
+                    exe.run(fluid.default_startup_program())
+                    r = exe.run(feed={'x': x_np}, fetch_list=[y])[0]
+                return r
+
+            x = np.random.randn(*shape).astype("float32")
+            y1 = compute_v1(x)
+            y2 = compute_v2(x)
+            self.assertTrue(np.allclose(y1, y2))
+
+
+if __name__ == '__main__':
+    unittest.main()

python/paddle/fluid/tests/unittests/test_layer_norm_op_v2.py

+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import os
+import unittest
+import numpy as np
+import paddle.fluid.core as core
+from paddle.fluid.op import Operator
+import paddle.fluid as fluid
+from op_test import OpTest, _set_use_system_allocator
+from paddle.fluid.framework import grad_var_name
+import paddle.fluid as fluid
+from paddle.fluid import Program, program_guard
+import paddle
+
+
+class TestDygraphLayerNormv2(unittest.TestCase):
+    def test_dygraph(self):
+        places = [fluid.CPUPlace()]
+        if core.is_compiled_with_cuda() and core.op_support_gpu("layer_norm"):
+            places.append(fluid.CUDAPlace(0))
+        for p in places:
+            shape = [4, 10, 4, 4]
+
+            def compute_v1(x):
+                with fluid.dygraph.guard(p):
+                    ln = fluid.dygraph.LayerNorm(shape[1:])
+                    y = ln(fluid.dygraph.to_variable(x))
+                return y.numpy()
+
+            def compute_v2(x):
+                with fluid.dygraph.guard(p):
+                    ln = paddle.nn.LayerNorm(shape[1:])
+                    y = ln(fluid.dygraph.to_variable(x))
+                return y.numpy()
+
+            x = np.random.randn(*shape).astype("float32")
+            y1 = compute_v1(x)
+            y2 = compute_v2(x)
+            self.assertTrue(np.allclose(y1, y2))
+
+    def test_static(self):
+        places = [fluid.CPUPlace()]
+        if core.is_compiled_with_cuda() and core.op_support_gpu("layer_norm"):
+            places.append(fluid.CUDAPlace(0))
+        for p in places:
+            exe = fluid.Executor(p)
+            shape = [4, 10, 16, 16]
+
+            def compute_v1(x_np):
+                with program_guard(Program(), Program()):
+                    ln = fluid.dygraph.LayerNorm(shape[1:])
+                    x = fluid.data(name='x', shape=x_np.shape, dtype=x_np.dtype)
+                    y = ln(x)
+                    exe.run(fluid.default_startup_program())
+                    r = exe.run(feed={'x': x_np}, fetch_list=[y])[0]
+                return r
+
+            def compute_v2(x_np):
+                with program_guard(Program(), Program()):
+                    ln = paddle.nn.LayerNorm(shape[1:])
+                    x = fluid.data(name='x', shape=x_np.shape, dtype=x_np.dtype)
+                    y = ln(x)
+                    exe.run(fluid.default_startup_program())
+                    r = exe.run(feed={'x': x_np}, fetch_list=[y])[0]
+                return r
+
+            x = np.random.randn(*shape).astype("float32")
+            y1 = compute_v1(x)
+            y2 = compute_v2(x)
+            self.assertTrue(np.allclose(y1, y2))
+
+
+if __name__ == '__main__':
+    unittest.main()

python/paddle/nn/__init__.py

@@ -128,6 +128,12 @@ from .layer.norm import GroupNorm  #DEFINE_ALIAS
 from .layer.norm import LayerNorm  #DEFINE_ALIAS
 from .layer.norm import SpectralNorm  #DEFINE_ALIAS
 from .layer.norm import InstanceNorm  #DEFINE_ALIAS
+from .layer.norm import InstanceNorm1d  #DEFINE_ALIAS
+from .layer.norm import InstanceNorm2d  #DEFINE_ALIAS
+from .layer.norm import InstanceNorm3d  #DEFINE_ALIAS
+from .layer.norm import BatchNorm1d  #DEFINE_ALIAS
+from .layer.norm import BatchNorm2d  #DEFINE_ALIAS
+from .layer.norm import BatchNorm3d  #DEFINE_ALIAS
 # from .layer.rnn import RNNCell        #DEFINE_ALIAS
 # from .layer.rnn import GRUCell        #DEFINE_ALIAS
 # from .layer.rnn import LSTMCell        #DEFINE_ALIAS

python/paddle/nn/functional/__init__.py

@@ -160,12 +160,12 @@ from .loss import square_error_cost  #DEFINE_ALIAS
 from .loss import ssd_loss  #DEFINE_ALIAS
 from .loss import teacher_student_sigmoid_loss  #DEFINE_ALIAS
 from .loss import ctc_loss  #DEFINE_ALIAS
-# from .norm import batch_norm        #DEFINE_ALIAS
 # from .norm import data_norm        #DEFINE_ALIAS
 # from .norm import group_norm        #DEFINE_ALIAS
-# from .norm import instance_norm        #DEFINE_ALIAS
 from .norm import l2_normalize  #DEFINE_ALIAS
-# from .norm import layer_norm        #DEFINE_ALIAS
+from .norm import batch_norm  #DEFINE_ALIAS
+from .norm import instance_norm  #DEFINE_ALIAS
+from .norm import layer_norm  #DEFINE_ALIAS
 from .norm import lrn  #DEFINE_ALIAS
 from .norm import normalize  #DEFINE_ALIAS
 # from .norm import spectral_norm        #DEFINE_ALIAS

python/paddle/nn/functional/norm.py

@@ -18,16 +18,19 @@ import paddle.fluid as fluid
 from ...fluid.data_feeder import check_variable_and_dtype, check_type
 from ...fluid.layer_helper import LayerHelper
 from ...fluid.framework import in_dygraph_mode, core
+from ...framework import create_parameter
 from ...fluid.layers import l2_normalize  #DEFINE_ALIAS
 from ...fluid.layers import lrn  #DEFINE_ALIAS
+from ...fluid.initializer import Constant
+from ...fluid.param_attr import ParamAttr
+from ...fluid import core, dygraph_utils
 
 __all__ = [
-    #       'batch_norm',
+    'batch_norm',
     #       'data_norm',
-    #       'group_norm',
-    #       'instance_norm',
+    'instance_norm',
     'l2_normalize',
-    #       'layer_norm',
+    'layer_norm',
     'lrn',
     'normalize',
     #       'spectral_norm'
@@ -110,3 +113,286 @@ def normalize(x, p=2, axis=1, epsilon=1e-12, name=None):
     eps = out.block.create_var(dtype=out.dtype)
     paddle.fill_constant([1], out.dtype, epsilon, out=eps)
     return paddle.elementwise_div(x, paddle.maximum(out, eps), name=name)
+
+
+def batch_norm(x,
+               running_mean,
+               running_var,
+               weight,
+               bias,
+               training=False,
+               momentum=0.9,
+               epsilon=1e-05,
+               data_format="NCHW",
+               name=None):
+    """
+    Applies Batch Normalization as described in the paper Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift .
+
+    nn.functional.batch_norm is uesd for nn.BatchNorm1d, nn.BatchNorm2d, nn.BatchNorm3d. Please use above API for BatchNorm.
+    
+    Parameters:
+        x(Tesnor): input value. It's data type should be float32, float64.
+        running_mean(Tensor): running mean.
+        running_var(Tensor): running variance.
+        weight(Tensor): The weight tensor of batch_norm, can not be None.
+        bias(Tensor): The bias tensor of batch_norm can not be None. 
+        epsilon(float, optional): The small value added to the variance to prevent division by zero. Default: 1e-5.
+        momentum(float, optional): The value used for the moving_mean and moving_var computation. Default: 0.9.
+        training(bool, optional): True means train mode which compute by batch data and track global mean and var during train period. False means inference mode which compute by global mean and var which calculated by train period. Defalut False.
+        data_format(str, optional): Specify the input data format, may be "NC", "NCL", "NCHW" or "NCDHW". Defalut "NCHW".
+        name(str, optional): Name for the BatchNorm, default is None. For more information, please refer to :ref:`api_guide_Name`..
+
+    Returns:
+        None
+
+    Examples:
+        .. code-block:: python
+
+          import paddle
+          import numpy as np
+
+          paddle.disable_static()
+          x = np.random.seed(123)
+          x = np.random.random(size=(2, 1, 2, 3)).astype('float32')
+          running_mean = np.random.random(size=1).astype('float32')
+          running_variance = np.random.random(size=1).astype('float32')
+          weight_data = np.random.random(size=1).astype('float32')
+          bias_data = np.random.random(size=1).astype('float32')
+          x = paddle.to_tensor(x)
+          rm = paddle.to_tensor(running_mean)
+          rv = paddle.to_tensor(running_variance)
+          w = paddle.to_tensor(weight_data)
+          b = paddle.to_tensor(bias_data)
+          batch_norm_out = paddle.nn.functional.batch_norm(x, rm, rv, w, b)
+          print batch_norm_out
+    """
+
+    assert len(x.shape) >= 2, "input dim must be larger than 1"
+
+    # we use not training means use_global_status, more details see nn._BatchNormBase
+    use_global_stats = not training
+    # input ad out must share the memory
+    mean_out = running_mean
+    variance_out = running_var
+
+    if in_dygraph_mode():
+        # for dygraph need tuple
+        attrs = ("momentum", momentum, "epsilon", epsilon, "data_layout",
+                 data_format, "use_mkldnn", False, "fuse_with_relu", False,
+                 "use_global_stats", use_global_stats)
+        batch_norm_out, _, _, _, _, _ = core.ops.batch_norm(
+            x, weight, bias, running_mean, running_var, mean_out, variance_out,
+            *attrs)
+
+        return dygraph_utils._append_activation_in_dygraph(
+            batch_norm_out, act=None)
+
+    check_variable_and_dtype(x, 'input', ['float16', 'float32', 'float64'],
+                             'BatchNorm')
+
+    # for static need dict
+    attrs = {
+        "momentum": momentum,
+        "epsilon": epsilon,
+        "data_layout": data_format,
+        "use_mkldnn": False,
+        "fuse_with_relu": False,
+        "use_global_stats": use_global_stats,
+    }
+
+    inputs = {
+        "X": [x],
+        "Scale": [weight],
+        "Bias": [bias],
+        "Mean": [running_mean],
+        "Variance": [running_var]
+    }
+
+    helper = LayerHelper('batch_norm', **locals())
+
+    dtype = x.dtype if x.dtype is not 'float16' else 'float32'
+    saved_mean = helper.create_variable_for_type_inference(
+        dtype=dtype, stop_gradient=True)
+    saved_variance = helper.create_variable_for_type_inference(
+        dtype=dtype, stop_gradient=True)
+    batch_norm_out = helper.create_variable_for_type_inference(dtype)
+
+    outputs = {
+        "Y": [batch_norm_out],
+        "MeanOut": [running_mean],
+        "VarianceOut": [running_var],
+        "SavedMean": [saved_mean],
+        "SavedVariance": [saved_variance]
+    }
+
+    helper.append_op(
+        type="batch_norm", inputs=inputs, outputs=outputs, attrs=attrs)
+
+    return helper.append_activation(batch_norm_out)
+
+
+def layer_norm(x,
+               normalized_shape,
+               weight=None,
+               bias=None,
+               epsilon=1e-05,
+               name=None):
+    """
+    see more detail in paddle.nn.LayerNorm
+    
+    Parameters:
+        x(Tensor): Input Tensor. It's data type should be float32, float64.
+        normalized_shape(int|list|tuple): Input shape from an expected input of
+            size :math:`[*, normalized_shape[0], normalized_shape[1], ..., normalized_shape[-1]]`.
+            If it is a single integer, this module will normalize over the last dimension
+            which is expected to be of that specific size.
+        epsilon(float, optional): The small value added to the variance to prevent
+            division by zero. Default: 1e-05.
+        weight(Tensor, optional): The weight tensor of batch_norm. Default: None.
+        bias(Tensor, optional): The bias tensor of batch_norm. Default: None.
+        name(str, optional): Name for the LayerNorm, default is None. For more information, please refer to :ref:`api_guide_Name`..
+
+    Returns:
+        None
+
+    Examples:
+
+        .. code-block:: python
+
+          import paddle
+          import numpy as np
+
+          paddle.disable_static()
+          np.random.seed(123)
+          x_data = np.random.random(size=(2, 2, 2, 3)).astype('float32')
+          x = paddle.to_tensor(x_data) 
+          layer_norm = paddle.nn.functional.layer_norm(x, x.shape[1:])
+          layer_norm_out = layer_norm(x)
+
+          print(layer_norm_out.numpy)
+    """
+    input_shape = list(x.shape)
+    input_ndim = len(input_shape)
+    normalized_ndim = len(normalized_shape)
+    begin_norm_axis = input_ndim - normalized_ndim
+    if input_ndim < normalized_ndim or input_shape[
+            begin_norm_axis:] != normalized_shape:
+        str_normalized_shape = str(normalized_shape)
+        raise ValueError('Given normalized_shape is ' + str_normalized_shape +
+                         ', expected input with shape [*, ' +
+                         str_normalized_shape[
+                             1:] + ', but got input shape ' + str(input_shape))
+
+    if in_dygraph_mode():
+        pre_act, _, _ = core.ops.layer_norm(x, weight, bias, 'epsilon', epsilon,
+                                            'begin_norm_axis', begin_norm_axis)
+        return dygraph_utils._append_activation_in_dygraph(pre_act, act=None)
+
+    check_variable_and_dtype(x, 'input', ['float32', 'float64'], 'LayerNorm')
+
+    inputs = dict()
+    inputs['X'] = [x]
+    if weight:
+        inputs['Scale'] = [weight]
+    if bias:
+        inputs['Bias'] = [bias]
+    attrs = {"epsilon": epsilon, "begin_norm_axis": begin_norm_axis}
+
+    # create output
+    helper = LayerHelper('layer_norm', **locals())
+    mean_out = helper.create_variable_for_type_inference(
+        dtype=x.type, stop_gradient=True)
+    variance_out = helper.create_variable_for_type_inference(
+        dtype=x.type, stop_gradient=True)
+    layer_norm_out = helper.create_variable_for_type_inference(x.type)
+
+    helper.append_op(
+        type="layer_norm",
+        inputs=inputs,
+        outputs={
+            "Y": layer_norm_out,
+            "Mean": mean_out,
+            "Variance": variance_out,
+        },
+        attrs={"epsilon": epsilon,
+               "begin_norm_axis": begin_norm_axis})
+
+    return helper.append_activation(layer_norm_out)
+
+
+def instance_norm(x,
+                  running_mean=None,
+                  running_var=None,
+                  weight=None,
+                  bias=None,
+                  use_input_stats=True,
+                  momentum=0.9,
+                  eps=1e-05,
+                  data_format="NCHW",
+                  name=None):
+    """
+    See more detail in nn.layer.InstanceNorm2d.
+
+    Parameters:
+        x(Tensor): Input Tensor. It's data type should be float32, float64.
+        running_mean(Tensor): running mean. Default None.
+        running_var(Tensor): running variance. Default None.
+        weight(Tensor, optional): The weight tensor of instance_norm. Default: None.
+        bias(Tensor, optional): The bias tensor of instance_norm. Default: None.
+        eps(float, optional): A value added to the denominator for numerical stability. Default is 1e-5.
+        momentum(float, optional): The value used for the moving_mean and moving_var computation. Default: 0.9.
+        use_input_stats(bool): Default True.
+        data_format(str, optional): Specify the input data format, may be "NC", "NCL", "NCHW" or "NCDHW". Defalut "NCHW".
+        name(str, optional): Name for the InstanceNorm, default is None. For more information, please refer to :ref:`api_guide_Name`..
+
+    Returns:
+        None.
+
+    Examples:
+
+        .. code-block:: python
+
+          import paddle
+          import numpy as np
+
+          paddle.disable_static()
+          np.random.seed(123)
+          x_data = np.random.random(size=(2, 2, 2, 3)).astype('float32')
+          x = paddle.to_tensor(x_data) 
+          instance_norm_out = paddle.nn.functional.instancenorm(x)
+
+          print(instance_norm_out.numpy)
+
+    """
+
+    if in_dygraph_mode():
+        out, _, _ = core.ops.instance_norm(x, weight, bias, "epsilon", eps,
+                                           "momentum", momentum, "data_format",
+                                           data_format)
+        return out
+
+    check_variable_and_dtype(x, 'input', ['float32', 'float64'], "InstanceNorm")
+
+    attrs = {"epsilon": eps, "momentum": momentum, "data_format": data_format}
+
+    if weight and bias:
+        inputs = {"X": [x], "Scale": [weight], "Bias": [bias]}
+    else:
+        inputs = {"X": [x]}
+
+    helper = LayerHelper('instance_norm', **locals())
+    saved_mean = helper.create_variable_for_type_inference(
+        dtype=x.dtype, stop_gradient=True)
+    saved_variance = helper.create_variable_for_type_inference(
+        dtype=x.dtype, stop_gradient=True)
+    instance_norm_out = helper.create_variable_for_type_inference(x.dtype)
+
+    outputs = {
+        "Y": [instance_norm_out],
+        "SavedMean": [saved_mean],
+        "SavedVariance": [saved_variance]
+    }
+
+    helper.append_op(
+        type="instance_norm", inputs=inputs, outputs=outputs, attrs=attrs)
+    return instance_norm_out