diff --git a/python/paddle/fluid/incubate/fleet/parameter_server/ir/trainer_pass.py b/python/paddle/fluid/incubate/fleet/parameter_server/ir/trainer_pass.py
index 5f327497047470dcafa93b7b07a51c63721b75ac..d4af3e2f8042a5dd83e5d4ba06cb3c89b352f8ed 100644
--- a/python/paddle/fluid/incubate/fleet/parameter_server/ir/trainer_pass.py
+++ b/python/paddle/fluid/incubate/fleet/parameter_server/ir/trainer_pass.py
@@ -527,7 +527,7 @@ def create_heter_program(program, config, heter_program, heter_ops,
     # This function mainly includes the following contents:
     # 1. For every heter block:
     #     a) copy heter device op from origin program
-    #     b) create variables which belong to heter op：
+    #     b) create variables which belong to heter op:
     #         -> if variable is persistable, clone it in global_scope
     #         -> if variable is temp, create it in heter block
     #     c) create communicate related op as follow:
diff --git a/python/paddle/fluid/tests/unittests/hccl_tools.py b/python/paddle/fluid/tests/unittests/hccl_tools.py
index 3ae8f38dc64bd1da6ab5a46ddd60b239b5461ad9..e3628ee5a4e9b4bf51950fccd424e17a69883eec 100644
--- a/python/paddle/fluid/tests/unittests/hccl_tools.py
+++ b/python/paddle/fluid/tests/unittests/hccl_tools.py
@@ -58,7 +58,7 @@ def parse_args():
         default="[0,8)",
         help="The number of the Ascend accelerators used. please note that the Ascend accelerators"
         "used must be continuous, such [0,4) means to use four chips "
-        "0，1，2，3; [0,1) means to use chip 0; The first four chips are"
+        "0,1,2,3; [0,1) means to use chip 0; The first four chips are"
         "a group, and the last four chips are a group. In addition to"
         "the [0,8) chips are allowed, other cross-group such as [3,6)"
         "are prohibited.")
diff --git a/python/paddle/nn/__init__.py b/python/paddle/nn/__init__.py
index 836d4008f7d0b50025f8816b003caa16c59792e3..d2f0063af0d220b7ae1f0945369bddb138860c06 100644
--- a/python/paddle/nn/__init__.py
+++ b/python/paddle/nn/__init__.py
@@ -15,148 +15,273 @@
 # TODO: import all neural network related api under this directory,
 # including layers, linear, conv, rnn etc.
 
-from .layer import norm
-from .functional import extension
-from .layer import common
-from .layer import rnn
-from .utils import weight_norm_hook
-
-from . import initializer
-
-__all__ = []
-__all__ += norm.__all__
-__all__ += extension.__all__
-__all__ += common.__all__
-__all__ += rnn.__all__
-__all__ += weight_norm_hook.__all__
-
-# TODO: define alias in nn directory
-from .clip import ClipGradByGlobalNorm  #DEFINE_ALIAS
-from .clip import ClipGradByNorm  #DEFINE_ALIAS
-from .clip import ClipGradByValue  #DEFINE_ALIAS
-# from .control_flow import cond  #DEFINE_ALIAS
-# from .control_flow import DynamicRNN        #DEFINE_ALIAS
-# from .control_flow import StaticRNN        #DEFINE_ALIAS
-# from .control_flow import while_loop  #DEFINE_ALIAS
-# from .control_flow import rnn        #DEFINE_ALIAS
-from .decode import BeamSearchDecoder  #DEFINE_ALIAS
-from .decode import dynamic_decode  #DEFINE_ALIAS
-# from .decode import Decoder        #DEFINE_ALIAS
-# from .decode import crf_decoding        #DEFINE_ALIAS
-# from .decode import ctc_greedy_decoder        #DEFINE_ALIAS
-# from .input import Input        #DEFINE_ALIAS
-from .layer.activation import ELU  #DEFINE_ALIAS
-from .layer.activation import GELU  #DEFINE_ALIAS
-from .layer.activation import Tanh  #DEFINE_ALIAS
-from .layer.activation import Hardshrink  #DEFINE_ALIAS
-from .layer.activation import Hardswish  #DEFINE_ALIAS
-from .layer.activation import Hardtanh  #DEFINE_ALIAS
-from .layer.activation import PReLU  #DEFINE_ALIAS
-from .layer.activation import ReLU  #DEFINE_ALIAS
-from .layer.activation import ReLU6  #DEFINE_ALIAS
-from .layer.activation import SELU  #DEFINE_ALIAS
-from .layer.activation import Silu  #DEFINE_ALIAS
-from .layer.activation import LeakyReLU  #DEFINE_ALIAS
-from .layer.activation import Sigmoid  #DEFINE_ALIAS
-from .layer.activation import Hardsigmoid  #DEFINE_ALIAS
-from .layer.activation import LogSigmoid  #DEFINE_ALIAS
-from .layer.activation import Softmax  #DEFINE_ALIAS
-from .layer.activation import Softplus  #DEFINE_ALIAS
-from .layer.activation import Softshrink  #DEFINE_ALIAS
-from .layer.activation import Softsign  #DEFINE_ALIAS
-from .layer.activation import Swish  #DEFINE_ALIAS
-from .layer.activation import Tanhshrink  #DEFINE_ALIAS
-from .layer.activation import ThresholdedReLU  #DEFINE_ALIAS
-from .layer.activation import LogSoftmax  #DEFINE_ALIAS
-from .layer.activation import Maxout  #DEFINE_ALIAS
-from .layer.common import Pad1D  #DEFINE_ALIAS
-from .layer.common import Pad2D  #DEFINE_ALIAS
-from .layer.common import Pad3D  #DEFINE_ALIAS
-from .layer.common import CosineSimilarity  #DEFINE_ALIAS
-from .layer.common import Embedding  #DEFINE_ALIAS
-from .layer.common import Linear  #DEFINE_ALIAS
-from .layer.common import Flatten  #DEFINE_ALIAS
-from .layer.common import Upsample  #DEFINE_ALIAS
-from .layer.common import UpsamplingNearest2D  #DEFINE_ALIAS
-from .layer.common import UpsamplingBilinear2D  #DEFINE_ALIAS
-from .layer.common import Bilinear  #DEFINE_ALIAS
-from .layer.common import Dropout  #DEFINE_ALIAS
-from .layer.common import Dropout2D  #DEFINE_ALIAS
-from .layer.common import Dropout3D  #DEFINE_ALIAS
-from .layer.common import AlphaDropout  #DEFINE_ALIAS
-from .layer.common import Unfold  #DEFINE_ALIAS
-
-from .layer.pooling import AvgPool1D  #DEFINE_ALIAS
-from .layer.pooling import AvgPool2D  #DEFINE_ALIAS
-from .layer.pooling import AvgPool3D  #DEFINE_ALIAS
-from .layer.pooling import MaxPool1D  #DEFINE_ALIAS
-from .layer.pooling import MaxPool2D  #DEFINE_ALIAS
-from .layer.pooling import MaxPool3D  #DEFINE_ALIAS
-from .layer.pooling import AdaptiveAvgPool1D  #DEFINE_ALIAS
-from .layer.pooling import AdaptiveAvgPool2D  #DEFINE_ALIAS
-from .layer.pooling import AdaptiveAvgPool3D  #DEFINE_ALIAS
-
-from .layer.pooling import AdaptiveMaxPool1D  #DEFINE_ALIAS
-from .layer.pooling import AdaptiveMaxPool2D  #DEFINE_ALIAS
-from .layer.pooling import AdaptiveMaxPool3D  #DEFINE_ALIAS
-from .layer.conv import Conv1D  #DEFINE_ALIAS
-from .layer.conv import Conv2D  #DEFINE_ALIAS
-from .layer.conv import Conv3D  #DEFINE_ALIAS
-from .layer.conv import Conv1DTranspose  #DEFINE_ALIAS
-from .layer.conv import Conv2DTranspose  #DEFINE_ALIAS
-from .layer.conv import Conv3DTranspose  #DEFINE_ALIAS
-# from .layer.conv import TreeConv        #DEFINE_ALIAS
-# from .layer.conv import Conv1D        #DEFINE_ALIAS
-from .layer.common import Linear
-# from .layer.loss import NCELoss        #DEFINE_ALIAS
-from .layer.loss import BCEWithLogitsLoss  #DEFINE_ALIAS
-from .layer.loss import CrossEntropyLoss  #DEFINE_ALIAS
-from .layer.loss import HSigmoidLoss  #DEFINE_ALIAS
-from .layer.loss import MSELoss  #DEFINE_ALIAS
-from .layer.loss import L1Loss  #DEFINE_ALIAS
-from .layer.loss import NLLLoss  #DEFINE_ALIAS
-from .layer.loss import BCELoss  #DEFINE_ALIAS
-from .layer.loss import KLDivLoss  #DEFINE_ALIAS
-from .layer.loss import MarginRankingLoss  #DEFINE_ALIAS
-from .layer.loss import CTCLoss  #DEFINE_ALIAS
-from .layer.loss import SmoothL1Loss  #DEFINE_ALIAS
-from .layer.norm import BatchNorm  #DEFINE_ALIAS
-from .layer.norm import SyncBatchNorm  #DEFINE_ALIAS
-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 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.norm import LocalResponseNorm  #DEFINE_ALIAS
-
-from .layer.rnn import RNNCellBase  #DEFINE_ALIAS
-from .layer.rnn import SimpleRNNCell  #DEFINE_ALIAS
-from .layer.rnn import LSTMCell  #DEFINE_ALIAS
-from .layer.rnn import GRUCell  #DEFINE_ALIAS
-from .layer.rnn import RNN  #DEFINE_ALIAS
-from .layer.rnn import BiRNN  #DEFINE_ALIAS
-from .layer.rnn import SimpleRNN  #DEFINE_ALIAS
-from .layer.rnn import LSTM  #DEFINE_ALIAS
-from .layer.rnn import GRU  #DEFINE_ALIAS
-
-from .layer.transformer import MultiHeadAttention
-from .layer.transformer import TransformerEncoderLayer
-from .layer.transformer import TransformerEncoder
-from .layer.transformer import TransformerDecoderLayer
-from .layer.transformer import TransformerDecoder
-from .layer.transformer import Transformer
-from .layer.distance import PairwiseDistance  #DEFINE_ALIAS
-
-from .layer.vision import PixelShuffle
-
-from .layer.container import LayerDict  #DEFINE_ALIAS
-
-from .layer import loss  #DEFINE_ALIAS
-from .layer import conv  #DEFINE_ALIAS
-from .layer import vision  #DEFINE_ALIAS
-from ..fluid.dygraph.layers import Layer  #DEFINE_ALIAS
-from ..fluid.dygraph.container import LayerList, ParameterList, Sequential  #DEFINE_ALIAS
+from .clip import ClipGradByGlobalNorm  # noqa: F401
+from .clip import ClipGradByNorm  # noqa: F401
+from .clip import ClipGradByValue  # noqa: F401
+from .decode import BeamSearchDecoder  # noqa: F401
+from .decode import dynamic_decode  # noqa: F401
+from .layer.activation import ELU  # noqa: F401
+from .layer.activation import GELU  # noqa: F401
+from .layer.activation import Tanh  # noqa: F401
+from .layer.activation import Hardshrink  # noqa: F401
+from .layer.activation import Hardswish  # noqa: F401
+from .layer.activation import Hardtanh  # noqa: F401
+from .layer.activation import PReLU  # noqa: F401
+from .layer.activation import ReLU  # noqa: F401
+from .layer.activation import ReLU6  # noqa: F401
+from .layer.activation import SELU  # noqa: F401
+from .layer.activation import Silu  # noqa: F401
+from .layer.activation import LeakyReLU  # noqa: F401
+from .layer.activation import Sigmoid  # noqa: F401
+from .layer.activation import Hardsigmoid  # noqa: F401
+from .layer.activation import LogSigmoid  # noqa: F401
+from .layer.activation import Softmax  # noqa: F401
+from .layer.activation import Softplus  # noqa: F401
+from .layer.activation import Softshrink  # noqa: F401
+from .layer.activation import Softsign  # noqa: F401
+from .layer.activation import Swish  # noqa: F401
+from .layer.activation import Tanhshrink  # noqa: F401
+from .layer.activation import ThresholdedReLU  # noqa: F401
+from .layer.activation import LogSoftmax  # noqa: F401
+from .layer.activation import Maxout  # noqa: F401
+from .layer.common import Pad1D  # noqa: F401
+from .layer.common import Pad2D  # noqa: F401
+from .layer.common import Pad3D  # noqa: F401
+from .layer.common import CosineSimilarity  # noqa: F401
+from .layer.common import Embedding  # noqa: F401
+from .layer.common import Linear  # noqa: F401
+from .layer.common import Flatten  # noqa: F401
+from .layer.common import Upsample  # noqa: F401
+from .layer.common import UpsamplingNearest2D  # noqa: F401
+from .layer.common import UpsamplingBilinear2D  # noqa: F401
+from .layer.common import Bilinear  # noqa: F401
+from .layer.common import Dropout  # noqa: F401
+from .layer.common import Dropout2D  # noqa: F401
+from .layer.common import Dropout3D  # noqa: F401
+from .layer.common import AlphaDropout  # noqa: F401
+from .layer.common import Unfold  # noqa: F401
+
+from .layer.pooling import AvgPool1D  # noqa: F401
+from .layer.pooling import AvgPool2D  # noqa: F401
+from .layer.pooling import AvgPool3D  # noqa: F401
+from .layer.pooling import MaxPool1D  # noqa: F401
+from .layer.pooling import MaxPool2D  # noqa: F401
+from .layer.pooling import MaxPool3D  # noqa: F401
+from .layer.pooling import AdaptiveAvgPool1D  # noqa: F401
+from .layer.pooling import AdaptiveAvgPool2D  # noqa: F401
+from .layer.pooling import AdaptiveAvgPool3D  # noqa: F401
+from .layer.pooling import AdaptiveMaxPool1D  # noqa: F401
+from .layer.pooling import AdaptiveMaxPool2D  # noqa: F401
+from .layer.pooling import AdaptiveMaxPool3D  # noqa: F401
+
+from .layer.conv import Conv1D  # noqa: F401
+from .layer.conv import Conv2D  # noqa: F401
+from .layer.conv import Conv3D  # noqa: F401
+from .layer.conv import Conv1DTranspose  # noqa: F401
+from .layer.conv import Conv2DTranspose  # noqa: F401
+from .layer.conv import Conv3DTranspose  # noqa: F401
+
+from .layer.loss import BCEWithLogitsLoss  # noqa: F401
+from .layer.loss import CrossEntropyLoss  # noqa: F401
+from .layer.loss import HSigmoidLoss  # noqa: F401
+from .layer.loss import MSELoss  # noqa: F401
+from .layer.loss import L1Loss  # noqa: F401
+from .layer.loss import NLLLoss  # noqa: F401
+from .layer.loss import BCELoss  # noqa: F401
+from .layer.loss import KLDivLoss  # noqa: F401
+from .layer.loss import MarginRankingLoss  # noqa: F401
+from .layer.loss import CTCLoss  # noqa: F401
+from .layer.loss import SmoothL1Loss  # noqa: F401
+from .layer.norm import BatchNorm  # noqa: F401
+from .layer.norm import SyncBatchNorm  # noqa: F401
+from .layer.norm import GroupNorm  # noqa: F401
+from .layer.norm import LayerNorm  # noqa: F401
+from .layer.norm import SpectralNorm  # noqa: F401
+from .layer.norm import InstanceNorm1D  # noqa: F401
+from .layer.norm import InstanceNorm2D  # noqa: F401
+from .layer.norm import InstanceNorm3D  # noqa: F401
+from .layer.norm import BatchNorm1D  # noqa: F401
+from .layer.norm import BatchNorm2D  # noqa: F401
+from .layer.norm import BatchNorm3D  # noqa: F401
+from .layer.norm import LocalResponseNorm  # noqa: F401
+
+from .layer.rnn import RNNCellBase  # noqa: F401
+from .layer.rnn import SimpleRNNCell  # noqa: F401
+from .layer.rnn import LSTMCell  # noqa: F401
+from .layer.rnn import GRUCell  # noqa: F401
+from .layer.rnn import RNN  # noqa: F401
+from .layer.rnn import BiRNN  # noqa: F401
+from .layer.rnn import SimpleRNN  # noqa: F401
+from .layer.rnn import LSTM  # noqa: F401
+from .layer.rnn import GRU  # noqa: F401
+
+from .layer.transformer import MultiHeadAttention  # noqa: F401
+from .layer.transformer import TransformerEncoderLayer  # noqa: F401
+from .layer.transformer import TransformerEncoder  # noqa: F401
+from .layer.transformer import TransformerDecoderLayer  # noqa: F401
+from .layer.transformer import TransformerDecoder  # noqa: F401
+from .layer.transformer import Transformer  # noqa: F401
+from .layer.distance import PairwiseDistance  # noqa: F401
+
+from .layer.vision import PixelShuffle  # noqa: F401
+from .layer.container import LayerDict  # noqa: F401
+
+# TODO: remove loss, keep it for too many used in unitests
+from .layer import loss  # noqa: F401
+from ..fluid.dygraph.layers import Layer  # noqa: F401
+from ..fluid.dygraph.container import LayerList  # noqa: F401
+from ..fluid.dygraph.container import ParameterList  # noqa: F401
+from ..fluid.dygraph.container import Sequential  # noqa: F401
+
+from . import utils  # noqa: F401
+from . import functional  # noqa: F401
+from . import initializer  # noqa: F401
+
+#TODO: remove 'diag_embed', 'remove_weight_norm', 'weight_norm' months later.
+import paddle.utils.deprecated as deprecated
+
+
+@deprecated(
+    since="2.0.0",
+    update_to="paddle.nn.funcitional.diag_embed",
+    reason="diag_embed in paddle.nn will removed in future")
+def diag_embed(*args):
+    '''
+        alias name of paddle.nn.functional.diag_embed
+    '''
+    return functional.diag_embed(*args)
+
+
+@deprecated(
+    since="2.0.0",
+    update_to="paddle.nn.utils.remove_weight_norm",
+    reason="remove_weight_norm in paddle.nn will removed in future")
+def remove_weight_norm(*args):
+    '''
+        alias name of paddle.nn.utils.remove_weight_norm
+    '''
+    return utils.remove_weight_norm(*args)
+
+
+@deprecated(
+    since="2.0.0",
+    update_to="paddle.nn.utils.weight_norm",
+    reason="weight_norm in paddle.nn will removed in future")
+def weight_norm(*args):
+    '''
+        alias name of paddle.nn.utils.weight_norm
+    '''
+    return utils.weight_norm(*args)
+
+
+__all__ = [     #noqa
+           'BatchNorm',
+           'GroupNorm',
+           'LayerNorm',
+           'SpectralNorm',
+           'BatchNorm1D',
+           'BatchNorm2D',
+           'BatchNorm3D',
+           'InstanceNorm1D',
+           'InstanceNorm2D',
+           'InstanceNorm3D',
+           'SyncBatchNorm',
+           'LocalResponseNorm',
+           'Embedding',
+           'Linear',
+           'Upsample',
+           'UpsamplingNearest2D',
+           'UpsamplingBilinear2D',
+           'Pad1D',
+           'Pad2D',
+           'Pad3D',
+           'CosineSimilarity',
+           'Dropout',
+           'Dropout2D',
+           'Dropout3D',
+           'Bilinear',
+           'AlphaDropout',
+           'Unfold'
+           'RNNCellBase',
+           'SimpleRNNCell',
+           'LSTMCell',
+           'GRUCell',
+           'RNN',
+           'BiRNN',
+           'SimpleRNN',
+           'LSTM',
+           'GRU',
+           'dynamic_decode',
+           'MultiHeadAttention',
+           'Maxout',
+           'Softsign',
+           'Transformer',
+           'MSELoss',
+           'LogSigmoid',
+           'BeamSearchDecoder',
+           'ClipGradByNorm',
+           'ReLU',
+           'PairwiseDistance',
+           'BCEWithLogitsLoss',
+           'SmoothL1Loss',
+           'MaxPool3D',
+           'AdaptiveMaxPool2D',
+           'Hardshrink',
+           'clip',
+           'Softplus',
+           'KLDivLoss',
+           'clip_by_norm',
+           'AvgPool2D',
+           'L1Loss',
+           'LeakyReLU',
+           'AvgPool1D',
+           'AdaptiveAvgPool3D',
+           'AdaptiveMaxPool3D',
+           'NLLLoss',
+           'Conv1D',
+           'Sequential',
+           'Hardswish',
+           'Conv1DTranspose',
+           'AdaptiveMaxPool1D',
+           'TransformerEncoder',
+           'Softmax',
+           'ParameterList',
+           'Conv2D',
+           'Softshrink',
+           'Hardtanh',
+           'TransformerDecoderLayer',
+           'CrossEntropyLoss',
+           'GELU',
+           'SELU',
+           'Silu',
+           'Conv2DTranspose',
+           'CTCLoss',
+           'ThresholdedReLU',
+           'AdaptiveAvgPool2D',
+           'MaxPool1D',
+           'Layer',
+           'TransformerDecoder',
+           'Conv3D',
+           'Tanh',
+           'Conv3DTranspose',
+           'Flatten',
+           'AdaptiveAvgPool1D',
+           'Tanhshrink',
+           'HSigmoidLoss',
+           'PReLU',
+           'TransformerEncoderLayer',
+           'AvgPool3D',
+           'MaxPool2D',
+           'MarginRankingLoss',
+           'LayerList',
+           'ClipGradByValue',
+           'BCELoss',
+           'Hardsigmoid',
+           'ClipGradByGlobalNorm',
+           'LogSoftmax',
+           'Sigmoid',
+           'Swish',
+           'PixelShuffle',
+           'ELU',
+           'ReLU6'
+]
diff --git a/python/paddle/nn/clip.py b/python/paddle/nn/clip.py
index 9180a883e835c367530ba449a7a2dccf158dedf5..70c49b4a53876db5705d63c3df501fbd374f0a29 100644
--- a/python/paddle/nn/clip.py
+++ b/python/paddle/nn/clip.py
@@ -13,8 +13,6 @@
 # limitations under the License.
 
 # TODO: define the functions to clip gradient of parameter  
-from ..fluid.clip import ClipGradByGlobalNorm  #DEFINE_ALIAS
-from ..fluid.clip import ClipGradByNorm  #DEFINE_ALIAS
-from ..fluid.clip import ClipGradByValue  #DEFINE_ALIAS
-
-__all__ = ['ClipGradByGlobalNorm', 'ClipGradByNorm', 'ClipGradByValue']
+from ..fluid.clip import ClipGradByGlobalNorm  # noqa: F401
+from ..fluid.clip import ClipGradByNorm  # noqa: F401
+from ..fluid.clip import ClipGradByValue  # noqa: F401
diff --git a/python/paddle/nn/decode.py b/python/paddle/nn/decode.py
index bba5aba0da9ad024823750c32f2e02bb22dfbbbb..3229f0b21a6699d1b7437955af31004538a791c9 100644
--- a/python/paddle/nn/decode.py
+++ b/python/paddle/nn/decode.py
@@ -12,10 +12,5 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from ..fluid.layers import BeamSearchDecoder  #DEFINE_ALIAS
-from ..fluid.layers import dynamic_decode  #DEFINE_ALIAS
-
-__all__ = [
-    'BeamSearchDecoder',
-    'dynamic_decode',
-]
+from ..fluid.layers import BeamSearchDecoder  # noqa: F401
+from ..fluid.layers import dynamic_decode  # noqa: F401
diff --git a/python/paddle/nn/functional/__init__.py b/python/paddle/nn/functional/__init__.py
index 98124be7288d008f573047d99c38a8b9b1236b2e..d4c17a27a61780b431916b2634585de035778ce8 100644
--- a/python/paddle/nn/functional/__init__.py
+++ b/python/paddle/nn/functional/__init__.py
@@ -14,211 +14,185 @@
 
 # TODO: import all neural network related api under this directory,
 # including layers, linear, conv, rnn etc.
-__all__ = []
 
-# TODO: define alias in functional directory
-from . import conv
-__all__ += conv.__all__
-from . import activation
-__all__ += activation.__all__
-from . import extension
-__all__ += extension.__all__
-from . import common
-__all__ += common.__all__
-from . import pooling
-__all__ += pooling.__all__
-from . import loss
-__all__ += loss.__all__
-from .activation import elu  #DEFINE_ALIAS
-from .activation import elu_  #DEFINE_ALIAS
-# from .activation import erf  #DEFINE_ALIAS
-from .activation import gelu  #DEFINE_ALIAS
-from .activation import hardshrink  #DEFINE_ALIAS
-from .activation import hardtanh  #DEFINE_ALIAS
-from .activation import hardsigmoid  #DEFINE_ALIAS
-from .activation import hardswish  #DEFINE_ALIAS
-from .activation import leaky_relu  #DEFINE_ALIAS
-from .activation import log_sigmoid  #DEFINE_ALIAS
-from .activation import maxout  #DEFINE_ALIAS
-from .activation import prelu  #DEFINE_ALIAS
-from .activation import relu  #DEFINE_ALIAS
-from .activation import relu_  #DEFINE_ALIAS
-from .activation import relu6  #DEFINE_ALIAS
-from .activation import selu  #DEFINE_ALIAS
-from .activation import sigmoid  #DEFINE_ALIAS
-from .activation import silu  #DEFINE_ALIAS
-# from .activation import soft_relu  #DEFINE_ALIAS
-from .activation import softmax  #DEFINE_ALIAS
-from .activation import softmax_  #DEFINE_ALIAS
-from .activation import softplus  #DEFINE_ALIAS
-from .activation import softshrink  #DEFINE_ALIAS
-from .activation import softsign  #DEFINE_ALIAS
-from .activation import swish  #DEFINE_ALIAS
-from .activation import tanh  #DEFINE_ALIAS
-from .activation import tanh_  #DEFINE_ALIAS
-from .activation import tanhshrink  #DEFINE_ALIAS
-from .activation import thresholded_relu  #DEFINE_ALIAS
-from .activation import log_softmax  #DEFINE_ALIAS
-from .activation import glu  #DEFINE_ALIAS
-from .common import dropout  #DEFINE_ALIAS
-from .common import dropout2d  #DEFINE_ALIAS
-from .common import dropout3d  #DEFINE_ALIAS
-from .common import alpha_dropout  #DEFINE_ALIAS
-# from .common import embedding        #DEFINE_ALIAS
-# from .common import fc  #DEFINE_ALIAS
-from .common import label_smooth
-# from .common import one_hot  #DEFINE_ALIAS
-from .common import pad  #DEFINE_ALIAS
-# from .common import pad_constant_like  #DEFINE_ALIAS
-# from .common import pad2d  #DEFINE_ALIAS
-from .common import cosine_similarity  #DEFINE_ALIAS
-from .common import unfold  #DEFINE_ALIAS
-# from .common import bilinear_tensor_product        #DEFINE_ALIAS
-from .common import interpolate  #DEFINE_ALIAS
-from .common import upsample  #DEFINE_ALIAS
-from .common import bilinear  #DEFINE_ALIAS
-from .conv import conv1d  #DEFINE_ALIAS
-from .conv import conv1d_transpose  #DEFINE_ALIAS
-from .common import linear  #DEFINE_ALIAS
-from .conv import conv2d  #DEFINE_ALIAS
-from .conv import conv2d_transpose  #DEFINE_ALIAS
-from .conv import conv3d  #DEFINE_ALIAS
-from .conv import conv3d_transpose  #DEFINE_ALIAS
-# from .extension import add_position_encoding  #DEFINE_ALIAS
-# from .extension import autoincreased_step_counter        #DEFINE_ALIAS
-# from .extension import continuous_value_model  #DEFINE_ALIAS
-# from .extension import filter_by_instag  #DEFINE_ALIAS
-# from .extension import linear_chain_crf        #DEFINE_ALIAS
-# from .extension import merge_selected_rows        #DEFINE_ALIAS
-# from .extension import multiclass_nms  #DEFINE_ALIAS
-# from .extension import polygon_box_transform  #DEFINE_ALIAS
-# from .extension import random_crop  #DEFINE_ALIAS
-# from .extension import rpn_target_assign  #DEFINE_ALIAS
-# from .extension import similarity_focus  #DEFINE_ALIAS
-# from .extension import target_assign  #DEFINE_ALIAS
-# from .extension import temporal_shift  #DEFINE_ALIAS
-# from .extension import warpctc  #DEFINE_ALIAS
-from .extension import diag_embed  #DEFINE_ALIAS
+from .activation import elu  # noqa: F401
+from .activation import elu_  # noqa: F401
+from .activation import gelu  # noqa: F401
+from .activation import hardshrink  # noqa: F401
+from .activation import hardtanh  # noqa: F401
+from .activation import hardsigmoid  # noqa: F401
+from .activation import hardswish  # noqa: F401
+from .activation import leaky_relu  # noqa: F401
+from .activation import log_sigmoid  # noqa: F401
+from .activation import maxout  # noqa: F401
+from .activation import prelu  # noqa: F401
+from .activation import relu  # noqa: F401
+from .activation import relu_  # noqa: F401
+from .activation import relu6  # noqa: F401
+from .activation import selu  # noqa: F401
+from .activation import sigmoid  # noqa: F401
+from .activation import silu  # noqa: F401
+from .activation import softmax  # noqa: F401
+from .activation import softmax_  # noqa: F401
+from .activation import softplus  # noqa: F401
+from .activation import softshrink  # noqa: F401
+from .activation import softsign  # noqa: F401
+from .activation import swish  # noqa: F401
+from .activation import tanh  # noqa: F401
+from .activation import tanh_  # noqa: F401
+from .activation import tanhshrink  # noqa: F401
+from .activation import thresholded_relu  # noqa: F401
+from .activation import log_softmax  # noqa: F401
+from .activation import glu  # noqa: F401
+from .common import dropout  # noqa: F401
+from .common import dropout2d  # noqa: F401
+from .common import dropout3d  # noqa: F401
+from .common import alpha_dropout  # noqa: F401
+from .common import label_smooth  # noqa: F401
+from .common import pad  # noqa: F401
+from .common import cosine_similarity  # noqa: F401
+from .common import unfold  # noqa: F401
+from .common import interpolate  # noqa: F401
+from .common import upsample  # noqa: F401
+from .common import bilinear  # noqa: F401
+from .conv import conv1d  # noqa: F401
+from .conv import conv1d_transpose  # noqa: F401
+from .common import linear  # noqa: F401
+from .conv import conv2d  # noqa: F401
+from .conv import conv2d_transpose  # noqa: F401
+from .conv import conv3d  # noqa: F401
+from .conv import conv3d_transpose  # noqa: F401
+from .extension import diag_embed  # noqa: F401
 from .extension import sequence_mask
-# from .lod import sequence_concat        #DEFINE_ALIAS
-# from .lod import sequence_conv        #DEFINE_ALIAS
-# from .lod import sequence_enumerate        #DEFINE_ALIAS
-# from .lod import sequence_expand_as        #DEFINE_ALIAS
-# from .lod import sequence_expand        #DEFINE_ALIAS
-# from .lod import sequence_first_step        #DEFINE_ALIAS
-# from .lod import sequence_last_step        #DEFINE_ALIAS
-# from .lod import sequence_mask        #DEFINE_ALIAS
-# from .lod import sequence_pad        #DEFINE_ALIAS
-# from .lod import sequence_pool        #DEFINE_ALIAS
-# from .lod import sequence_reshape        #DEFINE_ALIAS
-# from .lod import sequence_reverse        #DEFINE_ALIAS
-# from .lod import sequence_scatter        #DEFINE_ALIAS
-# from .lod import sequence_slice        #DEFINE_ALIAS
-# from .lod import sequence_softmax        #DEFINE_ALIAS
-# from .lod import sequence_unpad        #DEFINE_ALIAS
-# from .lod import array_length        #DEFINE_ALIAS
-# from .lod import array_read        #DEFINE_ALIAS
-# from .lod import array_write        #DEFINE_ALIAS
-# from .lod import create_array        #DEFINE_ALIAS
-# from .lod import hash  #DEFINE_ALIAS
-# from .lod import im2sequence        #DEFINE_ALIAS
-# from .lod import lod_append        #DEFINE_ALIAS
-# from .lod import lod_reset        #DEFINE_ALIAS
-# from .lod import reorder_lod_tensor_by_rank        #DEFINE_ALIAS
-# from .lod import tensor_array_to_tensor        #DEFINE_ALIAS
-# from .lod import dynamic_gru        #DEFINE_ALIAS
-# from .lod import dynamic_lstm        #DEFINE_ALIAS
-# from .lod import dynamic_lstmp        #DEFINE_ALIAS
-from .loss import binary_cross_entropy  #DEFINE_ALIAS
-from .loss import binary_cross_entropy_with_logits  #DEFINE_ALIAS
-# from .loss import bpr_loss  #DEFINE_ALIAS
-# from .loss import center_loss  #DEFINE_ALIAS
-#from .loss import cross_entropy  #DEFINE_ALIAS
-from .loss import cross_entropy  #DEFINE_ALIAS
-from .loss import dice_loss  #DEFINE_ALIAS
-from .loss import hsigmoid_loss  #DEFINE_ALIAS
-from .loss import kl_div  #DEFINE_ALIAS
-from .loss import l1_loss  #DEFINE_ALIAS
-from .loss import log_loss  #DEFINE_ALIAS
-from .loss import margin_ranking_loss  #DEFINE_ALIAS
-from .loss import mse_loss  #DEFINE_ALIAS
-from .loss import nll_loss  #DEFINE_ALIAS
-# from .loss import nce        #DEFINE_ALIAS
-from .loss import npair_loss  #DEFINE_ALIAS
-from .loss import sigmoid_focal_loss  #DEFINE_ALIAS
-# from .loss import smooth_l1  #DEFINE_ALIAS
-from .loss import smooth_l1_loss  #DEFINE_ALIAS
-from .loss import softmax_with_cross_entropy  #DEFINE_ALIAS
-from .loss import square_error_cost  #DEFINE_ALIAS
-# from .loss import teacher_student_sigmoid_loss  #DEFINE_ALIAS
-from .loss import ctc_loss  #DEFINE_ALIAS
-# from .norm import data_norm        #DEFINE_ALIAS
-# from .norm import group_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 local_response_norm  #DEFINE_ALIAS
-from .norm import normalize  #DEFINE_ALIAS
-# from .norm import spectral_norm        #DEFINE_ALIAS
-# from .pooling import pool2d  #DEFINE_ALIAS
-# from .pooling import pool3d  #DEFINE_ALIAS
-from .pooling import avg_pool1d  #DEFINE_ALIAS
-from .pooling import avg_pool2d  #DEFINE_ALIAS
-from .pooling import avg_pool3d  #DEFINE_ALIAS
-from .pooling import max_pool1d  #DEFINE_ALIAS
-from .pooling import max_pool2d  #DEFINE_ALIAS
-from .pooling import max_pool3d  #DEFINE_ALIAS
+from .loss import binary_cross_entropy  # noqa: F401
+from .loss import binary_cross_entropy_with_logits  # noqa: F401
+from .loss import cross_entropy  # noqa: F401
+from .loss import dice_loss  # noqa: F401
+from .loss import hsigmoid_loss  # noqa: F401
+from .loss import kl_div  # noqa: F401
+from .loss import l1_loss  # noqa: F401
+from .loss import log_loss  # noqa: F401
+from .loss import margin_ranking_loss  # noqa: F401
+from .loss import mse_loss  # noqa: F401
+from .loss import nll_loss  # noqa: F401
+from .loss import npair_loss  # noqa: F401
+from .loss import sigmoid_focal_loss  # noqa: F401
+from .loss import smooth_l1_loss  # noqa: F401
+from .loss import softmax_with_cross_entropy  # noqa: F401
+from .loss import square_error_cost  # noqa: F401
+from .loss import ctc_loss  # noqa: F401
+from .norm import batch_norm  # noqa: F401
+from .norm import instance_norm  # noqa: F401
+from .norm import layer_norm  # noqa: F401
+from .norm import local_response_norm  # noqa: F401
+from .norm import normalize  # noqa: F401
+from .pooling import avg_pool1d  # noqa: F401
+from .pooling import avg_pool2d  # noqa: F401
+from .pooling import avg_pool3d  # noqa: F401
+from .pooling import max_pool1d  # noqa: F401
+from .pooling import max_pool2d  # noqa: F401
+from .pooling import max_pool3d  # noqa: F401
 
-from .pooling import adaptive_max_pool1d  #DEFINE_ALIAS
-from .pooling import adaptive_max_pool2d  #DEFINE_ALIAS
-from .pooling import adaptive_max_pool3d  #DEFINE_ALIAS
-from .pooling import adaptive_avg_pool1d  #DEFINE_ALIAS
-from .pooling import adaptive_avg_pool2d  #DEFINE_ALIAS
-from .pooling import adaptive_avg_pool3d  #DEFINE_ALIAS
+from .pooling import adaptive_max_pool1d  # noqa: F401
+from .pooling import adaptive_max_pool2d  # noqa: F401
+from .pooling import adaptive_max_pool3d  # noqa: F401
+from .pooling import adaptive_avg_pool1d  # noqa: F401
+from .pooling import adaptive_avg_pool2d  # noqa: F401
+from .pooling import adaptive_avg_pool3d  # noqa: F401
 
-# from .rnn import rnn  #DEFINE_ALIAS
-# from .rnn import birnn  #DEFINE_ALIAS
-# from .rnn import gru_unit        #DEFINE_ALIAS
-# from .rnn import lstm        #DEFINE_ALIAS
-# from .rnn import lstm_unit        #DEFINE_ALIAS
-# from .vision import affine_channel  #DEFINE_ALIAS
-from .vision import affine_grid  #DEFINE_ALIAS
-# from .vision import anchor_generator  #DEFINE_ALIAS
-# from .vision import bipartite_match  #DEFINE_ALIAS
-# from .vision import box_clip  #DEFINE_ALIAS
-# from .vision import box_coder  #DEFINE_ALIAS
-# from .vision import box_decoder_and_assign  #DEFINE_ALIAS
-# from .vision import collect_fpn_proposals  #DEFINE_ALIAS
-# from .vision import deformable_conv  #DEFINE_ALIAS
-# from .vision import deformable_roi_pooling  #DEFINE_ALIAS
-# from .vision import density_prior_box  #DEFINE_ALIAS
-# from .vision import detection_output  #DEFINE_ALIAS
-# from .vision import distribute_fpn_proposals  #DEFINE_ALIAS
-# from .vision import fsp_matrix  #DEFINE_ALIAS
-# from .vision import generate_mask_labels  #DEFINE_ALIAS
-# from .vision import generate_proposal_labels  #DEFINE_ALIAS
-# from .vision import generate_proposals  #DEFINE_ALIAS
-from .vision import grid_sample  #DEFINE_ALIAS
-# from .vision import image_resize  #DEFINE_ALIAS
-# from .vision import image_resize_short  #DEFINE_ALIAS
-# from .vision import multi_box_head  #DEFINE_ALIAS
-from .vision import pixel_shuffle  #DEFINE_ALIAS
-# from .vision import prior_box  #DEFINE_ALIAS
-# from .vision import prroi_pool  #DEFINE_ALIAS
-# from .vision import psroi_pool  #DEFINE_ALIAS
-# from .vision import resize_bilinear  #DEFINE_ALIAS
-# from .vision import resize_nearest  #DEFINE_ALIAS
-# from .vision import resize_trilinear  #DEFINE_ALIAS
-# from .vision import retinanet_detection_output  #DEFINE_ALIAS
-# from .vision import retinanet_target_assign  #DEFINE_ALIAS
-# from .vision import roi_align  #DEFINE_ALIAS
-# from .vision import roi_perspective_transform  #DEFINE_ALIAS
-# from .vision import roi_pool  #DEFINE_ALIAS
-# from .vision import shuffle_channel  #DEFINE_ALIAS
-# from .vision import space_to_depth  #DEFINE_ALIAS
-# from .vision import yolo_box  #DEFINE_ALIAS
-# from .vision import yolov3_loss  #DEFINE_ALIAS
-from .input import one_hot  #DEFINE_ALIAS
-from .input import embedding  #DEFINE_ALIAS
-from ...fluid.layers import gather_tree
-from ...fluid.layers import temporal_shift
+from .vision import affine_grid  # noqa: F401
+from .vision import grid_sample  # noqa: F401
+from .vision import pixel_shuffle  # noqa: F401
+from .input import one_hot  # noqa: F401
+from .input import embedding  # noqa: F401
+from ...fluid.layers import gather_tree  # noqa: F401
+from ...fluid.layers import temporal_shift  # noqa: F401
+
+__all__ = [     #noqa
+           'conv1d',
+           'conv1d_transpose',
+           'conv2d',
+           'conv2d_transpose',
+           'conv3d',
+           'conv3d_transpose',
+           'elu',
+           'elu_',
+           'gelu',
+           'hardshrink',
+           'hardtanh',
+           'hardsigmoid',
+           'hardswish',
+           'leaky_relu',
+           'log_sigmoid',
+           'maxout',
+           'prelu',
+           'relu',
+           'relu_',
+           'relu6',
+           'selu',
+           'softmax',
+           'softmax_',
+           'softplus',
+           'softshrink',
+           'softsign',
+           'sigmoid',
+           'silu',
+           'swish',
+           'tanh',
+           'tanh_',
+           'tanhshrink',
+           'thresholded_relu',
+           'log_softmax',
+           'glu',
+           'diag_embed',
+           'sequence_mask',
+           'dropout',
+           'dropout2d',
+           'dropout3d',
+           'alpha_dropout',
+           'label_smooth',
+           'linear',
+           'pad',
+           'unfold',
+           'interpolate',
+           'upsample',
+           'bilinear',
+           'cosine_similarity',
+           'avg_pool1d',
+           'avg_pool2d',
+           'avg_pool3d',
+           'max_pool1d',
+           'max_pool2d',
+           'max_pool3d',
+           'adaptive_avg_pool1d',
+           'adaptive_avg_pool2d',
+           'adaptive_avg_pool3d',
+           'adaptive_max_pool1d',
+           'adaptive_max_pool2d',
+           'adaptive_max_pool3d',
+           'binary_cross_entropy',
+           'binary_cross_entropy_with_logits',
+           'cross_entropy',
+           'dice_loss',
+           'hsigmoid_loss',
+           'kl_div',
+           'l1_loss',
+           'log_loss',
+           'mse_loss',
+           'margin_ranking_loss',
+           'nll_loss',
+           'npair_loss',
+           'sigmoid_focal_loss',
+           'smooth_l1_loss',
+           'softmax_with_cross_entropy',
+           'square_error_cost',
+           'ctc_loss',
+           'affine_grid',
+           'grid_sample',
+           'local_response_norm',
+           'pixel_shuffle',
+           'embedding',
+           'gather_tree',
+           'one_hot',
+           'normalize'
+]
diff --git a/python/paddle/nn/functional/activation.py b/python/paddle/nn/functional/activation.py
index d74308dc9aa32a0960c59f2a2ff20e29741a4c88..cd8ee99baa23742b0c7c1ad5f20c7ce17e339790 100644
--- a/python/paddle/nn/functional/activation.py
+++ b/python/paddle/nn/functional/activation.py
@@ -12,53 +12,14 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-# TODO: define activation functions of neural network
-from ...fluid.layers import brelu  #DEFINE_ALIAS
-# from ...fluid.layers import erf  #DEFINE_ALIAS
-from ...fluid.layers import maxout  #DEFINE_ALIAS
-# from ...fluid.layers import soft_relu  #DEFINE_ALIAS
-from ...fluid.layers import swish  #DEFINE_ALIAS
-from ...fluid.layers import sigmoid  #DEFINE_ALIAS
-from ...tensor.math import tanh  #DEFINE_ALIAS
-from ...tensor.math import tanh_  #DEFINE_ALIAS
+from ...fluid.layers import sigmoid  # noqa: F401
+from ...tensor.math import tanh  # noqa: F401
+from ...tensor.math import tanh_  # noqa: F401
 
 from ...tensor.manipulation import _print_warning_in_static_mode
 from ...tensor.manipulation import chunk
 from ...tensor.math import multiply
 
-__all__ = [
-    'brelu',
-    'elu',
-    'elu_',
-    'gelu',
-    'hardshrink',
-    'hardtanh',
-    'hardsigmoid',
-    'hardswish',
-    'leaky_relu',
-    'log_sigmoid',
-    'maxout',
-    'prelu',
-    'relu',
-    'relu_',
-    'relu6',
-    'selu',
-    'softmax',
-    'softmax_',
-    'softplus',
-    'softshrink',
-    'softsign',
-    'sigmoid',
-    'silu'
-    'swish',
-    'tanh',
-    'tanh_',
-    'tanhshrink',
-    'thresholded_relu',
-    'log_softmax',
-    'glu',
-]
-
 import warnings
 from ...fluid.layer_helper import LayerHelper
 from ...fluid.framework import in_dygraph_mode, convert_np_dtype_to_dtype_
diff --git a/python/paddle/nn/functional/common.py b/python/paddle/nn/functional/common.py
index 1cc8ef6c39b15e348bc76cb014bde82b1c12e9bf..7379c7a5f67bd0a0a5661f5da41c46dca3bb16fb 100644
--- a/python/paddle/nn/functional/common.py
+++ b/python/paddle/nn/functional/common.py
@@ -20,44 +20,20 @@ from paddle.fluid.layers.tensor import Variable, fill_constant, zeros, concat
 from ...fluid.layers import core
 from ...fluid import dygraph_utils
 # TODO: define the common functions to build a neural network  
-# from ...fluid import one_hot  #DEFINE_ALIAS
-# from ...fluid.layers import pad2d  #DEFINE_ALIAS
-from ...fluid.layers import unfold  #DEFINE_ALIAS
-from ...fluid.layers import squeeze  #DEFINE_ALIAS
-from ...fluid.layers import unsqueeze  #DEFINE_ALIAS
+from ...fluid.layers import unfold  # noqa: F401
+from ...fluid.layers import squeeze
+from ...fluid.layers import unsqueeze
 from ...tensor import clip
 from ...tensor import sum
 from ...tensor import sqrt
-from ...tensor import sum  #DEFINE_ALIAS
-from ...tensor import sqrt  #DEFINE_ALIAS
 from ...fluid.data_feeder import check_variable_and_dtype, check_dtype
 from ...fluid.framework import Variable, in_dygraph_mode, _varbase_creator
 
-#from ...fluid.layers import fc  #DEFINE_ALIAS
-# from ...fluid.layers import pad_constant_like  #DEFINE_ALIAS
 from ...fluid.framework import in_dygraph_mode
 from ...fluid import core, dygraph_utils
 from ...fluid import core, layers
 from ...fluid.data_feeder import check_variable_and_dtype
 
-__all__ = [
-    'dropout',
-    'dropout2d',
-    'dropout3d',
-    'alpha_dropout',
-    #       'embedding',
-    #       'fc',
-    'label_smooth',
-    'linear',
-    'pad',
-    'unfold',
-    #       'bilinear_tensor_product',
-    'interpolate',
-    'upsample',
-    'bilinear',
-    'cosine_similarity',
-]
-
 
 def interpolate(x,
                 size=None,
diff --git a/python/paddle/nn/functional/conv.py b/python/paddle/nn/functional/conv.py
index a8d6a6cc38df2d8fec42f675bec814c8f7518d34..800c8204973723be905599b35d63d993bebbba94 100644
--- a/python/paddle/nn/functional/conv.py
+++ b/python/paddle/nn/functional/conv.py
@@ -13,15 +13,6 @@
 # limitations under the License.
 from __future__ import print_function
 
-__all__ = [
-    'conv1d',
-    'conv1d_transpose',
-    'conv2d',
-    'conv2d_transpose',
-    'conv3d',
-    'conv3d_transpose',
-]
-
 import numpy as np
 from ...device import get_cudnn_version
 from ...fluid.framework import Variable, in_dygraph_mode
diff --git a/python/paddle/nn/functional/extension.py b/python/paddle/nn/functional/extension.py
index b004d79a877e7edbccc2245e5056753ed6b14bfb..7900f903e7fd2f5bc298d18b28fd602f309238e9 100644
--- a/python/paddle/nn/functional/extension.py
+++ b/python/paddle/nn/functional/extension.py
@@ -14,8 +14,6 @@
 
 # TODO: define the extention functions
 
-__all__ = ['diag_embed', 'sequence_mask']
-
 import numpy as np
 from ...fluid.data_feeder import check_dtype
 from ...fluid.layer_helper import LayerHelper
diff --git a/python/paddle/nn/functional/input.py b/python/paddle/nn/functional/input.py
index b88a2b042ff48160c12aedcca6f12591c154cd0e..4fff9cda4be33abd5db3384837efc84c7a1063a7 100644
--- a/python/paddle/nn/functional/input.py
+++ b/python/paddle/nn/functional/input.py
@@ -19,8 +19,6 @@ from ...fluid.layer_helper import LayerHelper
 from ...fluid.layers import core
 from ...fluid.data_feeder import check_variable_and_dtype, check_dtype
 
-__all__ = ['one_hot', 'embedding']
-
 
 def one_hot(x, num_classes, name=None):
     """
diff --git a/python/paddle/nn/functional/loss.py b/python/paddle/nn/functional/loss.py
index ca0ad06532d27e6508cb3eabcfb82de70785f8df..bb2d8005f4e3186d0531ca2516e07d169f76afc5 100755
--- a/python/paddle/nn/functional/loss.py
+++ b/python/paddle/nn/functional/loss.py
@@ -24,14 +24,14 @@ import paddle
 import paddle.fluid as fluid
 from ...fluid.framework import core, in_dygraph_mode
 from ...fluid.layers.nn import _elementwise_op_in_dygraph
-from ...fluid.layers import dice_loss  #DEFINE_ALIAS
-from ...fluid.layers import log_loss  #DEFINE_ALIAS
-from ...fluid.layers import npair_loss  #DEFINE_ALIAS
+from ...fluid.layers import dice_loss  # noqa: F401
+from ...fluid.layers import log_loss  # noqa: F401
+from ...fluid.layers import npair_loss  # noqa: F401
 from ...fluid.layers import reshape
-from ...fluid.layers import softmax_with_cross_entropy as fluid_softmax_with_cross_entropy  #DEFINE_ALIAS
-from ...fluid.layers import square_error_cost  #DEFINE_ALIAS
+from ...fluid.layers import softmax_with_cross_entropy as fluid_softmax_with_cross_entropy
+from ...fluid.layers import square_error_cost  # noqa: F401
 
-from ...fluid.layers import edit_distance  #DEFINE_ALIAS
+from ...fluid.layers import edit_distance  # noqa: F401
 from ...fluid.layers import huber_loss
 from ...fluid.layer_helper import LayerHelper
 from ...fluid.framework import in_dygraph_mode
@@ -39,27 +39,6 @@ from ...fluid.framework import _varbase_creator
 from ...fluid.framework import Variable
 from paddle.utils import deprecated
 
-__all__ = [
-    'binary_cross_entropy',
-    'binary_cross_entropy_with_logits',
-    'cross_entropy',
-    'dice_loss',
-    'hsigmoid_loss',
-    'kl_div',
-    'l1_loss',
-    'log_loss',
-    'mse_loss',
-    'margin_ranking_loss',
-    #       'nce',
-    'nll_loss',
-    'npair_loss',
-    'sigmoid_focal_loss',
-    'smooth_l1_loss',
-    'softmax_with_cross_entropy',
-    'square_error_cost',
-    'ctc_loss',
-]
-
 
 def binary_cross_entropy(input, label, weight=None, reduction='mean',
                          name=None):
@@ -1312,7 +1291,7 @@ def cross_entropy(input,
             Indicate whether compute softmax before cross_entropy.
             Default is ``True``.
 
-        - **name** (str，optional)
+        - **name** (str, optional)
 
             The name of the operator. Default is ``None`` .
             For more information, please refer to :ref:`api_guide_Name` .
diff --git a/python/paddle/nn/functional/norm.py b/python/paddle/nn/functional/norm.py
index 73df03e3714c7926f267bbe6e105096964d8f3b0..dddc4c66d591ccbe86722492a571fd7d65570559 100644
--- a/python/paddle/nn/functional/norm.py
+++ b/python/paddle/nn/functional/norm.py
@@ -22,19 +22,8 @@ from ...framework import create_parameter
 from ...fluid.initializer import Constant
 from ...fluid.param_attr import ParamAttr
 from ...fluid import core, dygraph_utils
-
 import numbers
 
-__all__ = [
-    'batch_norm',
-    #       'data_norm',
-    'instance_norm',
-    'layer_norm',
-    'local_response_norm',
-    'normalize',
-    #       'spectral_norm'
-]
-
 
 def normalize(x, p=2, axis=1, epsilon=1e-12, name=None):
     r"""
diff --git a/python/paddle/nn/functional/pooling.py b/python/paddle/nn/functional/pooling.py
index 5f3642710ae0adfbdb53f7b5adc81c8b8395a924..27a66c629cafaa283346b07750392710d525a32c 100755
--- a/python/paddle/nn/functional/pooling.py
+++ b/python/paddle/nn/functional/pooling.py
@@ -18,21 +18,6 @@ from ...fluid.framework import in_dygraph_mode
 from ...fluid.layers import utils, LayerHelper, unsqueeze, squeeze
 from ...fluid.data_feeder import check_type, check_variable_and_dtype
 
-__all__ = [
-    'avg_pool1d',
-    'avg_pool2d',
-    'avg_pool3d',
-    'max_pool1d',
-    'max_pool2d',
-    'max_pool3d',
-    'adaptive_avg_pool1d',
-    'adaptive_avg_pool2d',
-    'adaptive_avg_pool3d',
-    'adaptive_max_pool1d',
-    'adaptive_max_pool2d',
-    'adaptive_max_pool3d',
-]
-
 
 def _is_list_or_tuple(input):
     return isinstance(input, (list, tuple))
diff --git a/python/paddle/nn/functional/vision.py b/python/paddle/nn/functional/vision.py
index 032d5b47eda077ab3a1cb9123d958568bf20617a..cb8a817023d22c243349f564bb2bd13f072da957 100644
--- a/python/paddle/nn/functional/vision.py
+++ b/python/paddle/nn/functional/vision.py
@@ -19,43 +19,6 @@ from ...fluid.data_feeder import check_variable_and_dtype
 from ...fluid import dygraph_utils
 import numpy as np
 
-# TODO: define specitial functions used in computer vision task  
-# from ...fluid.layers import affine_channel  #DEFINE_ALIAS
-# from ...fluid.layers import anchor_generator  #DEFINE_ALIAS
-# from ...fluid.layers import bipartite_match  #DEFINE_ALIAS
-# from ...fluid.layers import box_clip  #DEFINE_ALIAS
-# from ...fluid.layers import box_coder  #DEFINE_ALIAS
-# from ...fluid.layers import box_decoder_and_assign  #DEFINE_ALIAS
-# from ...fluid.layers import collect_fpn_proposals  #DEFINE_ALIAS
-# from ...fluid.layers import deformable_roi_pooling  #DEFINE_ALIAS
-# from ...fluid.layers import density_prior_box  #DEFINE_ALIAS
-# from ...fluid.layers import detection_output  #DEFINE_ALIAS
-# from ...fluid.layers import distribute_fpn_proposals  #DEFINE_ALIAS
-# from ...fluid.layers import generate_mask_labels  #DEFINE_ALIAS
-# from ...fluid.layers import generate_proposal_labels  #DEFINE_ALIAS
-# from ...fluid.layers import generate_proposals  #DEFINE_ALIAS
-# from ...fluid.layers import image_resize  #DEFINE_ALIAS
-# from ...fluid.layers import prior_box  #DEFINE_ALIAS
-# from ...fluid.layers import prroi_pool  #DEFINE_ALIAS
-# from ...fluid.layers import psroi_pool  #DEFINE_ALIAS
-# from ...fluid.layers import resize_bilinear  #DEFINE_ALIAS
-# from ...fluid.layers import resize_nearest  #DEFINE_ALIAS
-# from ...fluid.layers import resize_trilinear  #DEFINE_ALIAS
-# from ...fluid.layers import roi_align  #DEFINE_ALIAS
-# from ...fluid.layers import roi_pool  #DEFINE_ALIAS
-# from ...fluid.layers import space_to_depth  #DEFINE_ALIAS
-# from ...fluid.layers import yolo_box  #DEFINE_ALIAS
-# from ...fluid.layers import yolov3_loss  #DEFINE_ALIAS
-# from ...fluid.layers import fsp_matrix  #DEFINE_ALIAS
-# from ...fluid.layers import image_resize_short  #DEFINE_ALIAS
-# from ...fluid.layers import pixel_shuffle  #DEFINE_ALIAS
-# from ...fluid.layers import retinanet_detection_output  #DEFINE_ALIAS
-# from ...fluid.layers import retinanet_target_assign  #DEFINE_ALIAS
-# from ...fluid.layers import roi_perspective_transform  #DEFINE_ALIAS
-# from ...fluid.layers import shuffle_channel  #DEFINE_ALIAS
-
-__all__ = ['affine_grid', 'grid_sample', 'pixel_shuffle']
-
 
 def affine_grid(theta, out_shape, align_corners=True, name=None):
     """
diff --git a/python/paddle/nn/initializer/__init__.py b/python/paddle/nn/initializer/__init__.py
index c128a1b401b2d85c8076d773c441c2182b8327a4..03e91f80dd139ced85c8a0e011d776e80d43f5a8 100644
--- a/python/paddle/nn/initializer/__init__.py
+++ b/python/paddle/nn/initializer/__init__.py
@@ -13,36 +13,34 @@
 # limitations under the License.
 
 # TODO: define the initializers to create a Parameter in neural network
-from ...fluid.initializer import Bilinear  #DEFINE_ALIAS
-from ...fluid.initializer import set_global_initializer  #DEFINE_ALIAS
+from ...fluid.initializer import Bilinear  # noqa: F401
+from ...fluid.initializer import set_global_initializer  # noqa: F401
 
-from . import constant
-from .constant import Constant  #DEFINE_ALIAS
+from .constant import Constant  # noqa: F401
 
-from . import kaiming
-from .kaiming import KaimingNormal  #DEFINE_ALIAS
-from .kaiming import KaimingUniform  #DEFINE_ALIAS
+from .kaiming import KaimingNormal  # noqa: F401
+from .kaiming import KaimingUniform  # noqa: F401
 
-__all__ = ['Bilinear', 'set_global_initializer']
+from .xavier import XavierNormal  # noqa: F401
+from .xavier import XavierUniform  # noqa: F401
 
-__all__ += constant.__all__
-__all__ += kaiming.__all__
+from .assign import Assign  # noqa: F401
 
-from . import xavier
-from .xavier import XavierNormal  #DEFINE_ALIAS
-from .xavier import XavierUniform  #DEFINE_ALIAS
+from .normal import Normal  # noqa: F401
+from .normal import TruncatedNormal  # noqa: F401
 
-from . import assign
-from .assign import Assign  #DEFINE_ALIAS
+from .uniform import Uniform  # noqa: F401
 
-from . import normal
-from .normal import Normal  #DEFINE_ALIAS
-from .normal import TruncatedNormal  #DEFINE_ALIAS
-
-from . import uniform
-from .uniform import Uniform  #DEFINE_ALIAS
-
-__all__ += xavier.__all__
-__all__ += assign.__all__
-__all__ += normal.__all__
-__all__ += uniform.__all__
+__all__ = [     #noqa
+           'Bilinear',
+           'Constant',
+           'KaimingUniform',
+           'KaimingNormal',
+           'XavierNormal',
+           'XavierUniform',
+           'Assign',
+           'Normal',
+           'TruncatedNormal',
+           'Uniform',
+           'set_global_initializer'
+]
diff --git a/python/paddle/nn/initializer/assign.py b/python/paddle/nn/initializer/assign.py
index 94c4ddc1938823653db6ef78a823430928c724ff..642919f3540753f499190a16637ae67836f36d9b 100644
--- a/python/paddle/nn/initializer/assign.py
+++ b/python/paddle/nn/initializer/assign.py
@@ -19,8 +19,6 @@ from ...fluid.core import VarDesc
 from ...fluid.data_feeder import check_type
 from ...fluid.initializer import NumpyArrayInitializer
 
-__all__ = ['Assign']
-
 
 class Assign(NumpyArrayInitializer):
     """Init an parameter with a numpy array, list, or tensor.
diff --git a/python/paddle/nn/initializer/constant.py b/python/paddle/nn/initializer/constant.py
index 6d21ddae0d16b5003bc6766b4106dd937727c2b1..aec3e82aab62b7591d93b83fea137147672f820c 100644
--- a/python/paddle/nn/initializer/constant.py
+++ b/python/paddle/nn/initializer/constant.py
@@ -15,8 +15,6 @@
 # TODO: define the initializers of Constant in neural network
 from ...fluid.initializer import ConstantInitializer
 
-__all__ = ['Constant']
-
 
 class Constant(ConstantInitializer):
     """Implement the constant initializer.
diff --git a/python/paddle/nn/initializer/kaiming.py b/python/paddle/nn/initializer/kaiming.py
index 7e2b6f787f85316c9ad4c3bedf91eef3b19cd50d..712bffccda102fc5ca1edaf85c641850027a4f01 100644
--- a/python/paddle/nn/initializer/kaiming.py
+++ b/python/paddle/nn/initializer/kaiming.py
@@ -15,8 +15,6 @@
 # TODO: define the initializers of Kaiming functions in neural network
 from ...fluid.initializer import MSRAInitializer
 
-__all__ = ['KaimingUniform', 'KaimingNormal']
-
 
 class KaimingNormal(MSRAInitializer):
     r"""Implements the Kaiming Normal initializer
diff --git a/python/paddle/nn/initializer/normal.py b/python/paddle/nn/initializer/normal.py
index a572d0e2c9216040e3ffa7e1c02841ebc1fc33ae..c009df780054ed9b727479d4029d002a3b1a521a 100644
--- a/python/paddle/nn/initializer/normal.py
+++ b/python/paddle/nn/initializer/normal.py
@@ -15,8 +15,6 @@
 from ...fluid.initializer import NormalInitializer
 from ...fluid.initializer import TruncatedNormalInitializer
 
-__all__ = ['Normal', 'TruncatedNormal']
-
 
 class Normal(NormalInitializer):
     """The Random Normal (Gaussian) distribution initializer.
diff --git a/python/paddle/nn/initializer/uniform.py b/python/paddle/nn/initializer/uniform.py
index a5d7d34efcf664a5bd46d7d3f06e2c542a8b4ef9..e54a4d2187b8d7742ab2a877a9faee97abc33d72 100644
--- a/python/paddle/nn/initializer/uniform.py
+++ b/python/paddle/nn/initializer/uniform.py
@@ -14,8 +14,6 @@
 
 from ...fluid.initializer import UniformInitializer
 
-__all__ = ['Uniform']
-
 
 class Uniform(UniformInitializer):
     """The random uniform distribution initializer.
diff --git a/python/paddle/nn/initializer/xavier.py b/python/paddle/nn/initializer/xavier.py
index 821a6984753105162e878c879cd5b960d2aa80e1..01a4a8887b489ab3492603d365da09ae5eb7ca22 100644
--- a/python/paddle/nn/initializer/xavier.py
+++ b/python/paddle/nn/initializer/xavier.py
@@ -14,8 +14,6 @@
 
 from ...fluid.initializer import XavierInitializer
 
-__all__ = ['XavierNormal', 'XavierUniform']
-
 
 class XavierNormal(XavierInitializer):
     r"""
diff --git a/python/paddle/nn/layer/__init__.py b/python/paddle/nn/layer/__init__.py
index 17c4ca5c5d11d20490c25616176f71dcf46c389c..64f0391fb6533e817f48a5cd9c626751856c58f6 100644
--- a/python/paddle/nn/layer/__init__.py
+++ b/python/paddle/nn/layer/__init__.py
@@ -14,90 +14,70 @@
 
 # TODO: define activation functions of neural network
 
-from . import activation
-from . import loss
-from . import conv
-from . import activation
-from . import norm
-from . import rnn
-from . import vision
-from . import distance
-from . import transformer
-from . import container
+from . import rnn  # noqa: F401
+from . import transformer  # noqa: F401
+from . import container  # noqa: F401
 
-from .activation import *
-from .loss import *
-from .conv import *
-from .activation import *
-from .norm import *
-from .rnn import *
-from .vision import *
+from .activation import PReLU  # noqa: F401
+from .activation import ReLU  # noqa: F401
+from .activation import ReLU6  # noqa: F401
+from .activation import LeakyReLU  # noqa: F401
+from .activation import Sigmoid  # noqa: F401
+from .activation import Softmax  # noqa: F401
+from .activation import LogSoftmax  # noqa: F401
+from .common import Bilinear  # noqa: F401
+from .common import Pad1D  # noqa: F401
+from .common import Pad2D  # noqa: F401
+from .common import Pad3D  # noqa: F401
+from .common import CosineSimilarity  # noqa: F401
+from .common import Embedding  # noqa: F401
+from .common import Linear  # noqa: F401
+from .common import Flatten  # noqa: F401
+from .common import Upsample  # noqa: F401
+from .common import Dropout  # noqa: F401
+from .common import Dropout2D  # noqa: F401
+from .common import Dropout3D  # noqa: F401
+from .common import AlphaDropout  # noqa: F401
+from .common import Upsample  # noqa: F401
+from .common import UpsamplingBilinear2D  # noqa: F401
+from .common import UpsamplingNearest2D  # noqa: F401
+from .pooling import AvgPool1D  # noqa: F401
+from .pooling import AvgPool2D  # noqa: F401
+from .pooling import AvgPool3D  # noqa: F401
+from .pooling import MaxPool1D  # noqa: F401
+from .pooling import MaxPool2D  # noqa: F401
+from .pooling import MaxPool3D  # noqa: F401
+from .pooling import AdaptiveAvgPool1D  # noqa: F401
+from .pooling import AdaptiveAvgPool2D  # noqa: F401
+from .pooling import AdaptiveAvgPool3D  # noqa: F401
+from .pooling import AdaptiveMaxPool1D  # noqa: F401
+from .pooling import AdaptiveMaxPool2D  # noqa: F401
+from .pooling import AdaptiveMaxPool3D  # noqa: F401
+from .conv import Conv1D  # noqa: F401
+from .conv import Conv2D  # noqa: F401
+from .conv import Conv3D  # noqa: F401
+from .conv import Conv1DTranspose  # noqa: F401
+from .conv import Conv2DTranspose  # noqa: F401
+from .conv import Conv3DTranspose  # noqa: F401
+from .loss import BCEWithLogitsLoss  # noqa: F401
+from .loss import CrossEntropyLoss  # noqa: F401
+from .loss import MSELoss  # noqa: F401
+from .loss import L1Loss  # noqa: F401
+from .loss import NLLLoss  # noqa: F401
+from .loss import BCELoss  # noqa: F401
+from .loss import KLDivLoss  # noqa: F401
+from .loss import MarginRankingLoss  # noqa: F401
+from .loss import CTCLoss  # noqa: F401
+from .loss import SmoothL1Loss  # noqa: F401
+from .norm import BatchNorm1D  # noqa: F401
+from .norm import BatchNorm2D  # noqa: F401
+from .norm import BatchNorm3D  # noqa: F401
+from .norm import SyncBatchNorm  # noqa: F401
+from .norm import GroupNorm  # noqa: F401
+from .norm import LayerNorm  # noqa: F401
+from .norm import SpectralNorm  # noqa: F401
+from .norm import LocalResponseNorm  # noqa: F401
 
-from .transformer import *
-from .activation import PReLU  #DEFINE_ALIAS
-from .activation import ReLU  #DEFINE_ALIAS
-from .activation import LeakyReLU  #DEFINE_ALIAS
-from .activation import Sigmoid  #DEFINE_ALIAS
-from .activation import Softmax  #DEFINE_ALIAS
-from .activation import LogSoftmax  #DEFINE_ALIAS
-from .common import Bilinear  #DEFINE_ALIAS
-from .common import Pad1D  #DEFINE_ALIAS
-from .common import Pad2D  #DEFINE_ALIAS
-from .common import Pad3D  #DEFINE_ALIAS
-from .common import CosineSimilarity  #DEFINE_ALIAS
-from .common import Embedding  #DEFINE_ALIAS
-from .common import Linear  #DEFINE_ALIAS
-from .common import Flatten  #DEFINE_ALIAS
-from .common import Upsample  #DEFINE_ALIAS
-from .common import Dropout  #DEFINE_ALIAS
-from .common import Dropout2D  #DEFINE_ALIAS
-from .common import Dropout3D  #DEFINE_ALIAS
-from .common import AlphaDropout  #DEFINE_ALIAS
-from .common import Upsample  #DEFINE_ALIAS
-from .common import UpsamplingBilinear2D  #DEFINE_ALIAS
-from .common import UpsamplingNearest2D  #DEFINE_ALIAS
-from .pooling import AvgPool1D  #DEFINE_ALIAS
-from .pooling import AvgPool2D  #DEFINE_ALIAS
-from .pooling import AvgPool3D  #DEFINE_ALIAS
-from .pooling import MaxPool1D  #DEFINE_ALIAS
-from .pooling import MaxPool2D  #DEFINE_ALIAS
-from .pooling import MaxPool3D  #DEFINE_ALIAS
-from .pooling import AdaptiveAvgPool1D  #DEFINE_ALIAS
-from .pooling import AdaptiveAvgPool2D  #DEFINE_ALIAS
-from .pooling import AdaptiveAvgPool3D  #DEFINE_ALIAS
-from .pooling import AdaptiveMaxPool1D  #DEFINE_ALIAS
-from .pooling import AdaptiveMaxPool2D  #DEFINE_ALIAS
-from .pooling import AdaptiveMaxPool3D  #DEFINE_ALIAS
-from .conv import Conv1D  #DEFINE_ALIAS
-from .conv import Conv2D  #DEFINE_ALIAS
-from .conv import Conv3D  #DEFINE_ALIAS
-from .conv import Conv1DTranspose  #DEFINE_ALIAS
-from .conv import Conv2DTranspose  #DEFINE_ALIAS
-from .conv import Conv3DTranspose  #DEFINE_ALIAS
-# from .conv import TreeConv        #DEFINE_ALIAS
-# from .conv import Conv1D        #DEFINE_ALIAS
-# from .loss import NCELoss        #DEFINE_ALIAS
-from .loss import BCEWithLogitsLoss  #DEFINE_ALIAS
-from .loss import CrossEntropyLoss  #DEFINE_ALIAS
-from .loss import MSELoss  #DEFINE_ALIAS
-from .loss import L1Loss  #DEFINE_ALIAS
-from .loss import NLLLoss  #DEFINE_ALIAS
-from .loss import BCELoss  #DEFINE_ALIAS
-from .loss import KLDivLoss  #DEFINE_ALIAS
-from .loss import MarginRankingLoss  #DEFINE_ALIAS
-from .loss import CTCLoss  #DEFINE_ALIAS
-from .loss import SmoothL1Loss  #DEFINE_ALIAS
-from .norm import BatchNorm  #DEFINE_ALIAS
-from .norm import SyncBatchNorm  #DEFINE_ALIAS
-from .norm import GroupNorm  #DEFINE_ALIAS
-from .norm import LayerNorm  #DEFINE_ALIAS
-from .norm import SpectralNorm  #DEFINE_ALIAS
-#from .norm import InstanceNorm  #DEFINE_ALIAS
-from .norm import LocalResponseNorm  #DEFINE_ALIAS
-# from .rnn import RNNCell        #DEFINE_ALIAS
-# from .rnn import GRUCell        #DEFINE_ALIAS
-# from .rnn import LSTMCell        #DEFINE_ALIAS
-
-from .vision import PixelShuffle  #DEFINE_ALIAS
-from .distance import PairwiseDistance  #DEFINE_ALIAS
-from .container import LayerDict  #DEFINE_ALIAS
+from .vision import PixelShuffle  # noqa: F401
+from .distance import PairwiseDistance  # noqa: F401
+from .container import LayerDict  # noqa: F401
diff --git a/python/paddle/nn/layer/activation.py b/python/paddle/nn/layer/activation.py
index 2a9ae310615ce4bb8ed43a0fc014524fa583ccee..c6ce4588ea5dab1f8095b9db83d6fb0b4d503ebe 100644
--- a/python/paddle/nn/layer/activation.py
+++ b/python/paddle/nn/layer/activation.py
@@ -14,33 +14,6 @@
 
 # TODO: define activation functions of neural network
 
-__all__ = [
-    'ELU',
-    'GELU',
-    'Hardshrink',
-    'Hardswish',
-    'Tanh',
-    'Hardtanh',
-    'PReLU',
-    'ReLU',
-    'ReLU6',
-    'SELU',
-    'LeakyReLU',
-    'Sigmoid',
-    'Silu',
-    'Hardsigmoid',
-    'Softmax',
-    'Softplus',
-    'Softshrink',
-    'Softsign',
-    'Swish',
-    'Tanhshrink',
-    'ThresholdedReLU',
-    'LogSigmoid',
-    'LogSoftmax',
-    'Maxout',
-]
-
 from ...fluid.dygraph import layers
 from ...fluid import core
 from ...fluid.framework import in_dygraph_mode
diff --git a/python/paddle/nn/layer/common.py b/python/paddle/nn/layer/common.py
index 8c001793715e511c40557d0865fcc299a6a248a4..058507ba5dec36ac6dba964e662e5176418b18ed 100644
--- a/python/paddle/nn/layer/common.py
+++ b/python/paddle/nn/layer/common.py
@@ -14,30 +14,12 @@
 
 # TODO: define the common classes to build a neural network
 import paddle
-from ...fluid.dygraph import Flatten  #DEFINE_ALIAS
+from ...fluid.dygraph import Flatten  # noqa: F401
 from ...fluid.dygraph import layers
 from ...fluid.framework import in_dygraph_mode
 from .. import functional as F
 from ...fluid.framework import _dygraph_tracer
 
-__all__ = [
-    'Embedding',
-    'Linear',
-    'Upsample',
-    'Pad1D',
-    'Pad2D',
-    'Pad3D',
-    'UpsamplingNearest2D',
-    'UpsamplingBilinear2D',
-    'CosineSimilarity',
-    'Dropout',
-    'Dropout2D',
-    'Dropout3D',
-    'Bilinear',
-    'AlphaDropout',
-    'Unfold',
-]
-
 
 def _npairs(x, n):
     if isinstance(x, (paddle.Tensor, list)):
diff --git a/python/paddle/nn/layer/conv.py b/python/paddle/nn/layer/conv.py
index d6ba04dad04c796b110c2948e9cf5dcefae46afe..2360dc17cf17108f7bbb1b1620632002842a8269 100644
--- a/python/paddle/nn/layer/conv.py
+++ b/python/paddle/nn/layer/conv.py
@@ -14,15 +14,6 @@
 
 # TODO: define classes of convolutional neural network
 
-__all__ = [
-    'Conv1D',
-    'Conv2D',
-    'Conv3D',
-    'Conv1DTranspose',
-    'Conv2DTranspose',
-    'Conv3DTranspose',
-]
-
 import numpy as np
 
 from ...fluid import get_flags
diff --git a/python/paddle/nn/layer/distance.py b/python/paddle/nn/layer/distance.py
index 72e0a1b2d6d2009e0edb2674b13299460996c104..7eb0fc1fbb5755dbbc4f77493d2affb7247e61b7 100644
--- a/python/paddle/nn/layer/distance.py
+++ b/python/paddle/nn/layer/distance.py
@@ -12,8 +12,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-__all__ = ['PairwiseDistance']
-
 import numpy as np
 
 import paddle
diff --git a/python/paddle/nn/layer/loss.py b/python/paddle/nn/layer/loss.py
index 2dfb3acca68e16fb95efb012895bf935cee11440..356b22c632cf5199319e3d46ba764f04833632c3 100644
--- a/python/paddle/nn/layer/loss.py
+++ b/python/paddle/nn/layer/loss.py
@@ -21,20 +21,6 @@ import paddle
 from .. import functional as F
 from paddle.fluid.framework import core, in_dygraph_mode, _varbase_creator
 
-__all__ = [
-    'BCEWithLogitsLoss',
-    'CrossEntropyLoss',
-    'HSigmoidLoss',
-    'MSELoss',
-    'L1Loss',
-    'NLLLoss',
-    'BCELoss',
-    'KLDivLoss',
-    'MarginRankingLoss',
-    'CTCLoss',
-    'SmoothL1Loss',
-]
-
 
 class BCEWithLogitsLoss(fluid.dygraph.Layer):
     r"""
@@ -295,7 +281,7 @@ class CrossEntropyLoss(fluid.dygraph.Layer):
             Indicate whether compute softmax before cross_entropy.
             Default is ``True``.
 
-        - **name** (str，optional)
+        - **name** (str, optional)
 
             The name of the operator. Default is ``None`` .
             For more information, please refer to :ref:`api_guide_Name` .
@@ -318,7 +304,7 @@ class CrossEntropyLoss(fluid.dygraph.Layer):
 
         - **label** (Tensor)
 
-            1. If soft_label=False，the shape is 
+            1. If soft_label=False, the shape is 
             :math:`[N_1, N_2, ..., N_k]` or :math:`[N_1, N_2, ..., N_k, 1]`, k >= 1.
             the data type is int32, int64, float32, float64, where each value is [0, C-1].
 
diff --git a/python/paddle/nn/layer/norm.py b/python/paddle/nn/layer/norm.py
index 0b0b2bf7b9b272ff5a41647b5d711b0069b2892a..970d68e8263432105349e9dfc699d0a0b2ea1115 100644
--- a/python/paddle/nn/layer/norm.py
+++ b/python/paddle/nn/layer/norm.py
@@ -28,13 +28,10 @@
 # TODO: define normalization api  
 
 import six
-#from ...fluid.dygraph.nn import InstanceNorm
 
-from ...fluid.dygraph import BatchNorm  #DEFINE_ALIAS
-#from ...fluid.dygraph import GroupNorm  #DEFINE_ALIAS
+from ...fluid.dygraph import BatchNorm  # noqa: F401
 
-#from ...fluid.dygraph import LayerNorm  #DEFINE_ALIAS
-from ...fluid.dygraph import SpectralNorm  #DEFINE_ALIAS
+from ...fluid.dygraph import SpectralNorm  # noqa: F401
 
 from ...fluid.dygraph import layers
 from ...framework import get_default_dtype, set_default_dtype
@@ -53,12 +50,6 @@ import warnings
 from ...fluid.dygraph.base import no_grad
 from .. import functional as F
 
-__all__ = [
-    'BatchNorm', 'GroupNorm', 'LayerNorm', 'SpectralNorm', 'BatchNorm1D',
-    'BatchNorm2D', 'BatchNorm3D', 'InstanceNorm1D', 'InstanceNorm2D',
-    'InstanceNorm3D', 'SyncBatchNorm', 'LocalResponseNorm'
-]
-
 
 class _InstanceNormBase(layers.Layer):
     """
diff --git a/python/paddle/nn/layer/pooling.py b/python/paddle/nn/layer/pooling.py
index cdb87a1cb3920748127032aef31f7710bdfc5641..5916fd7c69eb0be56dbeef2feaa6d4f8bb969b15 100755
--- a/python/paddle/nn/layer/pooling.py
+++ b/python/paddle/nn/layer/pooling.py
@@ -16,21 +16,6 @@ from ...fluid.dygraph import layers
 from ...fluid.layer_helper import LayerHelper
 from .. import functional as F
 
-__all__ = [
-    'AvgPool1D',
-    'AvgPool2D',
-    'AvgPool3D',
-    'MaxPool1D',
-    'MaxPool2D',
-    'MaxPool3D',
-    'AdaptiveAvgPool1D',
-    'AdaptiveAvgPool2D',
-    'AdaptiveAvgPool3D',
-    'AdaptiveMaxPool1D',
-    'AdaptiveMaxPool2D',
-    'AdaptiveMaxPool3D',
-]
-
 
 class AvgPool1D(layers.Layer):
     r"""
diff --git a/python/paddle/nn/layer/rnn.py b/python/paddle/nn/layer/rnn.py
index 964cfa74ebf0881cb7ceeed4c0e37d64f5005f98..a7539b5b09571c843247e1fce61d170aa71d91a0 100644
--- a/python/paddle/nn/layer/rnn.py
+++ b/python/paddle/nn/layer/rnn.py
@@ -33,18 +33,6 @@ from paddle.fluid.layers import utils
 from paddle.fluid.layers.utils import map_structure, flatten, pack_sequence_as
 from paddle.fluid.data_feeder import convert_dtype
 
-__all__ = [
-    'RNNCellBase',
-    'SimpleRNNCell',
-    'LSTMCell',
-    'GRUCell',
-    'RNN',
-    'BiRNN',
-    'SimpleRNN',
-    'LSTM',
-    'GRU',
-]
-
 
 def split_states(states, bidirectional=False, state_components=1):
     r"""
diff --git a/python/paddle/nn/layer/transformer.py b/python/paddle/nn/layer/transformer.py
index fe70a99ffb518fa17595c8a9a1817adcec457493..752870f3d0a28bd0e8db29138f6a5a901076515b 100644
--- a/python/paddle/nn/layer/transformer.py
+++ b/python/paddle/nn/layer/transformer.py
@@ -13,14 +13,6 @@
 # limitations under the License.
 
 # TODO: define the classes of Transformer neural network
-__all__ = [
-    'MultiHeadAttention',
-    'TransformerEncoderLayer',
-    'TransformerEncoder',
-    'TransformerDecoderLayer',
-    'TransformerDecoder',
-    'Transformer',
-]
 
 import copy
 import collections
diff --git a/python/paddle/nn/layer/vision.py b/python/paddle/nn/layer/vision.py
index d9c948a848a939c0427c14aee793e2c9c439c47b..e66e122be5259a8861b22d06258f34e1af89b6c9 100644
--- a/python/paddle/nn/layer/vision.py
+++ b/python/paddle/nn/layer/vision.py
@@ -17,8 +17,6 @@
 from ...fluid.dygraph import layers
 from .. import functional
 
-__all__ = ['PixelShuffle']
-
 
 class PixelShuffle(layers.Layer):
     """
diff --git a/python/paddle/nn/utils/__init__.py b/python/paddle/nn/utils/__init__.py
index 6562ac35e1e3180db671f90188f1304f07864189..bf2573d2cbc2d3a5229411abb857dd9d23031889 100644
--- a/python/paddle/nn/utils/__init__.py
+++ b/python/paddle/nn/utils/__init__.py
@@ -12,5 +12,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from . import weight_norm_hook
-from .weight_norm_hook import weight_norm, remove_weight_norm
+from .weight_norm_hook import weight_norm, remove_weight_norm  # noqa: F401
+
+__all__ = [  #noqa
+    'weight_norm', 'remove_weight_norm'
+]
diff --git a/python/paddle/nn/utils/weight_norm_hook.py b/python/paddle/nn/utils/weight_norm_hook.py
index fdf7a1b5bb2e2dc7e5e729a15c76fcbbb32ca12d..23df38ca08c45a4190a56e26fb47637eb349c407 100755
--- a/python/paddle/nn/utils/weight_norm_hook.py
+++ b/python/paddle/nn/utils/weight_norm_hook.py
@@ -19,8 +19,6 @@ from ...fluid import layers as F
 from ...fluid.layer_helper import LayerHelper
 from ...fluid.data_feeder import check_variable_and_dtype
 
-__all__ = ['weight_norm', 'remove_weight_norm']
-
 
 def l2_norm(x, axis, epsilon=1e-12, name=None):
     if len(x.shape) == 1:
diff --git a/python/paddle/utils/deprecated.py b/python/paddle/utils/deprecated.py
index daa2826ca360f120cea5f0fd0afecc8dc40b0b7e..a46f1ae3a2c2e3a255a1e7fefec25907c8c48e4c 100755
--- a/python/paddle/utils/deprecated.py
+++ b/python/paddle/utils/deprecated.py
@@ -83,13 +83,14 @@ def deprecated(update_to="", since="", reason=""):
                2. since version is empty, in this case, API is deprecated in all versions.
                3. current version is newer than since version.
             """
-            msg = "\033[93mWarning %s \033[0m" % (msg)
+            warningmsg = "\033[93mWarning %s \033[0m" % (msg)
             v_current = [int(i) for i in paddle.__version__.split(".")]
             v_current += [0] * (4 - len(v_current))
             v_since = [int(i) for i in _since.split(".")]
             v_since += [0] * (4 - len(v_since))
             if paddle.__version__ == "0.0.0" or _since == "" or v_current >= v_since:
-                warnings.warn(msg, category=DeprecationWarning, stacklevel=2)
+                warnings.warn(
+                    warningmsg, category=DeprecationWarning, stacklevel=2)
 
             return func(*args, **kwargs)