diff --git a/python/paddle/__init__.py b/python/paddle/__init__.py
index 7bac330376c44fb9632258b81ccb00255ab33a7c..e4cca3d459c4ce1a4c13162a18ea627b012954f4 100755
--- a/python/paddle/__init__.py
+++ b/python/paddle/__init__.py
@@ -21,8 +21,7 @@ except ImportError:
      import paddle from the source directory; please install paddlepaddle*.whl firstly.'''
                      )
 
-import paddle.batch
-batch = batch.batch
+from .batch import batch  # noqa: F401
 from .fluid import monkey_patch_variable
 from .fluid.dygraph import monkey_patch_math_varbase
 monkey_patch_variable()
@@ -135,7 +134,6 @@ from .tensor.manipulation import squeeze  # noqa: F401
 from .tensor.manipulation import squeeze_  # noqa: F401
 from .tensor.manipulation import stack  # noqa: F401
 from .tensor.manipulation import strided_slice  # noqa: F401
-from .tensor.manipulation import transpose  # noqa: F401
 from .tensor.manipulation import unique  # noqa: F401
 from .tensor.manipulation import unsqueeze  # noqa: F401
 from .tensor.manipulation import unsqueeze_  # noqa: F401
@@ -191,7 +189,6 @@ from .tensor.math import floor_mod  # noqa: F401
 from .tensor.math import multiply  # noqa: F401
 from .tensor.math import add  # noqa: F401
 from .tensor.math import subtract  # noqa: F401
-from .tensor.math import atan  # noqa: F401
 from .tensor.math import logsumexp  # noqa: F401
 from .tensor.math import inverse  # noqa: F401
 from .tensor.math import log1p  # noqa: F401
@@ -244,9 +241,8 @@ from .framework import save  # noqa: F401
 from .framework import load  # noqa: F401
 from .framework import DataParallel  # noqa: F401
 
-from .framework import set_default_dtype  #DEFINE_ALIAS
-from .framework import get_default_dtype  #DEFINE_ALIAS
-from .framework import set_grad_enabled  #DEFINE_ALIAS
+from .framework import set_default_dtype  # noqa: F401
+from .framework import get_default_dtype  # noqa: F401
 
 from .tensor.search import index_sample  # noqa: F401
 from .tensor.stat import mean  # noqa: F401
@@ -281,7 +277,7 @@ import paddle.vision  # noqa: F401
 from .tensor.random import check_shape  # noqa: F401
 disable_static()
 
-__all__ = [     #noqa
+__all__ = [  # noqa
            'dtype',
            'uint8',
            'int8',
@@ -323,7 +319,6 @@ __all__ = [     #noqa
            'cos',
            'tan',
            'mean',
-           'XPUPlace',
            'mv',
            'in_dynamic_mode',
            'min',
@@ -360,7 +355,6 @@ __all__ = [     #noqa
            'to_tensor',
            'gather_nd',
            'isinf',
-           'set_device',
            'uniform',
            'floor_divide',
            'remainder',
@@ -384,8 +378,6 @@ __all__ = [     #noqa
            'rand',
            'less_equal',
            'triu',
-           'is_compiled_with_cuda',
-           'is_compiled_with_rocm',
            'sin',
            'dist',
            'unbind',
@@ -414,8 +406,6 @@ __all__ = [     #noqa
            'bernoulli',
            'summary',
            'sinh',
-           'is_compiled_with_xpu',
-           'is_compiled_with_npu',
            'round',
            'DataParallel',
            'argmin',
@@ -437,7 +427,6 @@ __all__ = [     #noqa
            'not_equal',
            'sum',
            'tile',
-           'get_device',
            'greater_equal',
            'isfinite',
            'create_parameter',
@@ -470,7 +459,6 @@ __all__ = [     #noqa
            'scatter_nd',
            'set_default_dtype',
            'expand_as',
-           'get_cudnn_version',
            'stack',
            'sqrt',
            'cholesky',
@@ -484,7 +472,6 @@ __all__ = [     #noqa
            'logical_not',
            'add_n',
            'minimum',
-           'ComplexTensor',
            'scatter',
            'scatter_',
            'floor',
@@ -493,5 +480,6 @@ __all__ = [     #noqa
            'log2',
            'log10',
            'concat',
-           'check_shape'
+           'check_shape',
+           'standard_normal'
 ]
diff --git a/python/paddle/amp/__init__.py b/python/paddle/amp/__init__.py
index 32587938512c44df82cfa7353ac45a6cc3094186..64992752b2e8d8fcc90308b3f61c6a55abd01bc5 100644
--- a/python/paddle/amp/__init__.py
+++ b/python/paddle/amp/__init__.py
@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from .auto_cast import auto_cast
-from .grad_scaler import GradScaler
+from .auto_cast import auto_cast  # noqa: F401
+from .grad_scaler import GradScaler  # noqa: F401
 
 __all__ = ['auto_cast', 'GradScaler']
diff --git a/python/paddle/amp/auto_cast.py b/python/paddle/amp/auto_cast.py
index b83f81b27d1a0745c7a2f3339bc3939eb2f19490..974f718c2d4e2319c2f74783e285a4eb9365c80e 100644
--- a/python/paddle/amp/auto_cast.py
+++ b/python/paddle/amp/auto_cast.py
@@ -14,7 +14,7 @@
 
 from paddle.fluid.dygraph.amp import amp_guard
 
-__all__ = ['auto_cast']
+__all__ = []
 
 
 def auto_cast(enable=True, custom_white_list=None, custom_black_list=None):
diff --git a/python/paddle/amp/grad_scaler.py b/python/paddle/amp/grad_scaler.py
index 72a67a92c495863aba62bdaa93811e59780ed846..770b660a9e11ff4fad06deec7b2f4bbbdf1964a8 100644
--- a/python/paddle/amp/grad_scaler.py
+++ b/python/paddle/amp/grad_scaler.py
@@ -14,7 +14,7 @@
 
 from paddle.fluid.dygraph.amp import AmpScaler
 
-__all__ = ['GradScaler']
+__all__ = []
 
 
 class GradScaler(AmpScaler):
diff --git a/python/paddle/autograd/__init__.py b/python/paddle/autograd/__init__.py
index 71110e95817879fa55bcfa98293139a29b79997a..569619f065a051d071eb8be6b8d8f63049b20d2f 100644
--- a/python/paddle/autograd/__init__.py
+++ b/python/paddle/autograd/__init__.py
@@ -12,10 +12,9 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from ..fluid.dygraph.base import grad  #DEFINE_ALIAS
-
-from . import backward_mode
-from .backward_mode import backward
-from .py_layer import PyLayer, PyLayerContext
+from ..fluid.dygraph.base import grad  # noqa: F401
+from . import backward_mode  # noqa: F401
+from .backward_mode import backward  # noqa: F401
+from .py_layer import PyLayer, PyLayerContext  # noqa: F401
 
 __all__ = ['grad', 'backward', 'PyLayer', 'PyLayerContext']
diff --git a/python/paddle/autograd/backward_mode.py b/python/paddle/autograd/backward_mode.py
index 96e4336abaa6fa9ca5f23a56c551b8002c347888..6efbe777d537cadbe07a3bf21d807799e1227439 100644
--- a/python/paddle/autograd/backward_mode.py
+++ b/python/paddle/autograd/backward_mode.py
@@ -15,7 +15,7 @@
 from paddle.fluid import core
 from paddle.fluid import framework
 import paddle
-__all__ = ['backward']
+__all__ = []
 
 
 @framework.dygraph_only
diff --git a/python/paddle/autograd/py_layer.py b/python/paddle/autograd/py_layer.py
index 35e2cd24391775c6e9144d555e68ab12295385b6..5a22d22151a1cd12b68fc3672faec965f399d5fd 100644
--- a/python/paddle/autograd/py_layer.py
+++ b/python/paddle/autograd/py_layer.py
@@ -15,7 +15,7 @@
 import paddle
 from paddle.fluid.framework import dygraph_only
 from paddle.fluid import core
-__all__ = ['PyLayer', 'PyLayerContext']
+__all__ = []
 
 
 class PyLayerContext(object):
diff --git a/python/paddle/batch.py b/python/paddle/batch.py
index f6d2d8eb288744acab6d1c2f9d2a9db9a3087f58..f787f603f7e3ae0a6aa205596add48d192f54451 100644
--- a/python/paddle/batch.py
+++ b/python/paddle/batch.py
@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-__all__ = ['batch']
+__all__ = []
 
 
 def batch(reader, batch_size, drop_last=False):
@@ -35,11 +35,11 @@ def batch(reader, batch_size, drop_last=False):
     Examples:
         .. code-block:: python
            
-            import paddle.fluid as fluid
+            import paddle
             def reader():
                 for i in range(10):
                     yield i
-            batch_reader = fluid.io.batch(reader, batch_size=2)
+            batch_reader = paddle.batch(reader, batch_size=2)
             
             for data in batch_reader():
                 print(data)
@@ -60,7 +60,7 @@ def batch(reader, batch_size, drop_last=False):
             if len(b) == batch_size:
                 yield b
                 b = []
-        if drop_last == False and len(b) != 0:
+        if drop_last is False and len(b) != 0:
             yield b
 
     # Batch size check
diff --git a/python/paddle/compat.py b/python/paddle/compat.py
index 7c753815c5ccd32cfa65668b71b22e19b381b2b6..886a787623ed18cd4b7a56f7ac6661f5f7bc6800 100644
--- a/python/paddle/compat.py
+++ b/python/paddle/compat.py
@@ -15,18 +15,11 @@
 import six
 import math
 
-__all__ = [
-    'long_type',
-    'to_text',
-    'to_bytes',
-    'round',
-    'floor_division',
-    'get_exception_message',
-]
+__all__ = []
 
 if six.PY2:
     int_type = int
-    long_type = long
+    long_type = long  # noqa: F821
 else:
     int_type = int
     long_type = int
diff --git a/python/paddle/device.py b/python/paddle/device.py
index 85b813a7f51b56ec953df55a88e1e8288399ce1b..93e439ecf0aa420f178d646047a1b72c17189a65 100644
--- a/python/paddle/device.py
+++ b/python/paddle/device.py
@@ -18,21 +18,16 @@ import os
 from paddle.fluid import core
 from paddle.fluid import framework
 from paddle.fluid.dygraph.parallel import ParallelEnv
-from paddle.fluid.framework import is_compiled_with_cuda  #DEFINE_ALIAS
-from paddle.fluid.framework import is_compiled_with_rocm  #DEFINE_ALIAS
+from paddle.fluid.framework import is_compiled_with_cuda  # noqa: F401
+from paddle.fluid.framework import is_compiled_with_rocm  # noqa: F401
 
-__all__ = [
+
+__all__ = [  # npqa
     'get_cudnn_version',
     'set_device',
     'get_device',
     'XPUPlace',
-    'is_compiled_with_xpu'
-    #            'cpu_places',
-    #            'CPUPlace',
-    #            'cuda_pinned_places',
-    #            'cuda_places',
-    #            'CUDAPinnedPlace',
-    #            'CUDAPlace',
+    'is_compiled_with_xpu',
     'is_compiled_with_cuda',
     'is_compiled_with_rocm',
     'is_compiled_with_npu'
@@ -68,7 +63,7 @@ def is_compiled_with_xpu():
         .. code-block:: python
 
             import paddle
-            support_xpu = paddle.device.is_compiled_with_xpu()
+            support_xpu = paddle.is_compiled_with_xpu()
     """
     return core.is_compiled_with_xpu()
 
@@ -82,9 +77,10 @@ def XPUPlace(dev_id):
 
     Examples:
         .. code-block:: python
-
+            # required: xpu
+            
             import paddle
-            place = paddle.device.XPUPlace(0)
+            place = paddle.XPUPlace(0)
     """
     return core.XPUPlace(dev_id)
 
@@ -127,15 +123,13 @@ def _convert_to_place(device):
         place = core.CPUPlace()
     elif lower_device == 'gpu':
         if not core.is_compiled_with_cuda():
-            raise ValueError(
-                "The device should not be 'gpu', " \
-                "since PaddlePaddle is not compiled with CUDA")
+            raise ValueError("The device should not be 'gpu', "
+                             "since PaddlePaddle is not compiled with CUDA")
         place = core.CUDAPlace(ParallelEnv().dev_id)
     elif lower_device == 'xpu':
         if not core.is_compiled_with_xpu():
-            raise ValueError(
-                "The device should not be 'xpu', " \
-                "since PaddlePaddle is not compiled with XPU")
+            raise ValueError("The device should not be 'xpu', "
+                             "since PaddlePaddle is not compiled with XPU")
         selected_xpus = os.getenv("FLAGS_selected_xpus", "0").split(",")
         device_id = int(selected_xpus[0])
         place = core.XPUPlace(device_id)
@@ -149,7 +143,7 @@ def _convert_to_place(device):
         if avaliable_gpu_device:
             if not core.is_compiled_with_cuda():
                 raise ValueError(
-                    "The device should not be {}, since PaddlePaddle is " \
+                    "The device should not be {}, since PaddlePaddle is "
                     "not compiled with CUDA".format(avaliable_gpu_device))
             device_info_list = device.split(':', 1)
             device_id = device_info_list[1]
@@ -158,7 +152,7 @@ def _convert_to_place(device):
         if avaliable_xpu_device:
             if not core.is_compiled_with_xpu():
                 raise ValueError(
-                    "The device should not be {}, since PaddlePaddle is " \
+                    "The device should not be {}, since PaddlePaddle is "
                     "not compiled with XPU".format(avaliable_xpu_device))
             device_info_list = device.split(':', 1)
             device_id = device_info_list[1]
diff --git a/python/paddle/distributed/parallel.py b/python/paddle/distributed/parallel.py
index bc042e722947a0f0293f655a5d58f7823c0d0d03..efe747408428a68772726c28af469b975836511e 100644
--- a/python/paddle/distributed/parallel.py
+++ b/python/paddle/distributed/parallel.py
@@ -29,9 +29,7 @@ from paddle.fluid.dygraph import parallel_helper
 from paddle.fluid.dygraph.parallel import ParallelEnv
 from paddle.distributed.fleet.base.private_helper_function import wait_server_ready  # noqa: F401
 
-__all__ = [  #noqa
-    "init_parallel_env"
-]
+__all__ = []
 
 ParallelStrategy = core.ParallelStrategy
 
@@ -152,7 +150,6 @@ def init_parallel_env():
     init_gloo = int(os.getenv("PADDLE_WITH_GLOO", "0"))
     if init_gloo:
         ep_rank_0 = parallel_env.trainer_endpoints[0].split(":")
-        ep_rank = parallel_env.trainer_endpoints[parallel_env.rank].split(":")
         manager = Manager()
         # glboal dict to store status
         http_server_d = manager.dict()
diff --git a/python/paddle/incubate/__init__.py b/python/paddle/incubate/__init__.py
index 03e5a88624086b8781a1d8bee4437d9a17c98f76..22769053b1ac97508027910db28d91c526adead3 100644
--- a/python/paddle/incubate/__init__.py
+++ b/python/paddle/incubate/__init__.py
@@ -12,10 +12,11 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from . import optimizer
-from . import checkpoint
-from ..fluid.layer_helper import LayerHelper
+from .optimizer import LookAhead  # noqa: F401
+from .optimizer import ModelAverage  # noqa: F401
+from .checkpoint import auto_checkpoint  # noqa: F401
+from ..fluid.layer_helper import LayerHelper  # noqa: F401
 
-__all__ = []
-__all__ += optimizer.__all__
-__all__ += checkpoint.__all__
+__all__ = [  # noqa
+    'LookAhead', 'ModelAverage'
+]
diff --git a/python/paddle/incubate/checkpoint/__init__.py b/python/paddle/incubate/checkpoint/__init__.py
index 7ddd256df747981019e3afb0bb1dd839cf3ea550..79e6259de0275410664b9bfb2c34c33e21c5d529 100644
--- a/python/paddle/incubate/checkpoint/__init__.py
+++ b/python/paddle/incubate/checkpoint/__init__.py
@@ -12,6 +12,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from ...fluid.incubate.checkpoint import auto_checkpoint
+from ...fluid.incubate.checkpoint import auto_checkpoint  # noqa: F401
 
-__all__ = ["auto_checkpoint"]
+__all__ = []
diff --git a/python/paddle/incubate/optimizer/__init__.py b/python/paddle/incubate/optimizer/__init__.py
index 4a3889d0ee1a905a534f33909b4241f5c91be2f5..d966d187f288ac0865109cf361dd310328792aaf 100644
--- a/python/paddle/incubate/optimizer/__init__.py
+++ b/python/paddle/incubate/optimizer/__init__.py
@@ -12,7 +12,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from .lookahead import LookAhead
-from .modelaverage import ModelAverage
+from .lookahead import LookAhead  # noqa: F401
+from .modelaverage import ModelAverage  # noqa: F401
 
-__all__ = ['LookAhead', 'ModelAverage']
+__all__ = []
diff --git a/python/paddle/incubate/optimizer/lookahead.py b/python/paddle/incubate/optimizer/lookahead.py
index f90d520a5dfe8adc524ef20b7489ea008fb9c51a..720a84a24f0aa65c833939844c53e871b4e0680b 100644
--- a/python/paddle/incubate/optimizer/lookahead.py
+++ b/python/paddle/incubate/optimizer/lookahead.py
@@ -20,7 +20,7 @@ import paddle
 import numpy as np
 from paddle.fluid.dygraph import base as imperative_base
 
-__all__ = ["LookAhead"]
+__all__ = []
 
 
 class LookAhead(Optimizer):
@@ -99,7 +99,7 @@ class LookAhead(Optimizer):
             layer = LinearNet()
             loss_fn = nn.CrossEntropyLoss()
             optimizer = paddle.optimizer.SGD(learning_rate=0.1, parameters=layer.parameters())
-            lookahead = paddle.incubate.optimizer.LookAhead(optimizer, alpha=0.2, k=5)
+            lookahead = paddle.incubate.LookAhead(optimizer, alpha=0.2, k=5)
 
             # create data loader
             dataset = RandomDataset(BATCH_NUM * BATCH_SIZE)
@@ -163,7 +163,7 @@ class LookAhead(Optimizer):
                 out = linear(inp)
                 loss = paddle.mean(out)
                 sgd = paddle.optimizer.SGD(learning_rate=0.1,parameters=linear.parameters())
-                lookahead = paddle.incubate.optimizer.LookAhead(sgd, alpha=0.2, k=5)
+                lookahead = paddle.incubate.LookAhead(sgd, alpha=0.2, k=5)
                 loss.backward()
                 lookahead.step()
                 lookahead.clear_grad()
@@ -274,7 +274,7 @@ class LookAhead(Optimizer):
                 out = linear(inp)
                 loss = paddle.mean(out)
                 sgd = paddle.optimizer.SGD(learning_rate=0.1,parameters=linear.parameters())
-                lookahead = paddle.incubate.optimizer.LookAhead(sgd, alpha=0.2, k=5)
+                lookahead = paddle.incubate.LookAhead(sgd, alpha=0.2, k=5)
                 loss.backward()
                 lookahead.minimize(loss)
                 lookahead.clear_grad()
@@ -282,9 +282,6 @@ class LookAhead(Optimizer):
         """
         assert isinstance(loss, Variable), "The loss should be an Tensor."
 
-        parameter_list = parameters if parameters \
-            else self._parameter_list
-
         # Apply inner optimizer to the main_program
         optimize_ops, params_grads = self.inner_optimizer.minimize(
             loss,
diff --git a/python/paddle/incubate/optimizer/modelaverage.py b/python/paddle/incubate/optimizer/modelaverage.py
index 8afcaf9207e7cc84d143356b4f5efb74a175f2bd..8ffc3bdac62d040ccd45fe9768fdf566e784dcc4 100644
--- a/python/paddle/incubate/optimizer/modelaverage.py
+++ b/python/paddle/incubate/optimizer/modelaverage.py
@@ -21,7 +21,7 @@ import numpy as np
 from paddle.fluid.dygraph import base as imperative_base
 from paddle.fluid.wrapped_decorator import signature_safe_contextmanager
 
-__all__ = ["ModelAverage"]
+__all__ = []
 
 
 class ModelAverage(Optimizer):
@@ -129,7 +129,7 @@ class ModelAverage(Optimizer):
         layer = LinearNet()
         loss_fn = nn.CrossEntropyLoss()
         optimizer = opt.Momentum(learning_rate=0.2, momentum=0.1, parameters=layer.parameters())
-        model_average = paddle.incubate.optimizer.ModelAverage(0.15,
+        model_average = paddle.incubate.ModelAverage(0.15,
                                                     parameters=layer.parameters(),
                                                     min_average_window=2,
                                                     max_average_window=10)
@@ -313,7 +313,7 @@ class ModelAverage(Optimizer):
                 sgd = paddle.optimizer.SGD(learning_rate=0.1,parameters=linear.parameters())
                 sgd.minimize(loss)
 
-                modelaverage = paddle.incubate.optimizer.ModelAverage(0.15,
+                modelaverage = paddle.incubate.ModelAverage(0.15,
                                                             parameters=linear.parameters(),
                                                             min_average_window=2,
                                                             max_average_window=4)
@@ -345,7 +345,7 @@ class ModelAverage(Optimizer):
                 out = linear(inp)
                 loss = paddle.mean(out)
                 sgd = paddle.optimizer.SGD(learning_rate=0.1,parameters=linear.parameters())
-                modelaverage = paddle.incubate.optimizer.ModelAverage(0.15,
+                modelaverage = paddle.incubate.ModelAverage(0.15,
                                                             parameters=linear.parameters(),
                                                             min_average_window=2,
                                                             max_average_window=4)
@@ -395,7 +395,7 @@ class ModelAverage(Optimizer):
 
                 sgd = paddle.optimizer.SGD(learning_rate=0.1,parameters=linear.parameters())
 
-                modelaverage = paddle.incubate.optimizer.ModelAverage(0.15,
+                modelaverage = paddle.incubate.ModelAverage(0.15,
                                                             parameters=linear.parameters(),
                                                             min_average_window=2,
                                                             max_average_window=4)
@@ -415,7 +415,6 @@ class ModelAverage(Optimizer):
                                                         param)
                 old_num_accumulates = self._get_accumulator(
                     'old_num_accumulates', param)
-                num_updates = self._get_accumulator('num_updates', param)
                 sum_1 = self._get_accumulator('sum_1', param)
                 sum_2 = self._get_accumulator('sum_2', param)
                 sum_3 = self._get_accumulator('sum_3', param)
@@ -467,7 +466,7 @@ class ModelAverage(Optimizer):
 
                 sgd = paddle.optimizer.SGD(learning_rate=0.1,parameters=linear.parameters())
 
-                modelaverage = paddle.incubate.optimizer.ModelAverage(0.15,
+                modelaverage = paddle.incubate.ModelAverage(0.15,
                                                             parameters=linear.parameters(),
                                                             min_average_window=2,
                                                             max_average_window=4)
@@ -506,17 +505,15 @@ class ModelAverage(Optimizer):
             self._get_accumulator('num_accumulates', param))
         old_num_accumulates = block._clone_variable(
             self._get_accumulator('old_num_accumulates', param))
-        num_updates = block._clone_variable(
-            self._get_accumulator('num_updates', param))
         # backup param value to grad
         layers.assign(input=param, output=grad)
         # param = (sum_1 + sum_2 + sum_3) / (num_accumulates + old_num_accumulates)
         tmp = layers.sum(x=[num_accumulates, old_num_accumulates])
         sum = layers.sum(x=[sum_1, sum_2, sum_3])
         tmp = layers.cast(
-            x=tmp, dtype='float32' if self._dtype == None else self._dtype)
+            x=tmp, dtype='float32' if self._dtype is None else self._dtype)
         sum = layers.cast(
-            x=sum, dtype='float32' if self._dtype == None else self._dtype)
+            x=sum, dtype='float32' if self._dtype is None else self._dtype)
         layers.ops._elementwise_div(x=sum, y=tmp, out=param)
 
     def _add_average_restore_op(self, block, param):
diff --git a/python/paddle/inference/__init__.py b/python/paddle/inference/__init__.py
index c388301ec3408e436eacb2567e8e529d0bbc03bb..4e172039716628a157c8324c17ff2d4be3666349 100644
--- a/python/paddle/inference/__init__.py
+++ b/python/paddle/inference/__init__.py
@@ -12,5 +12,26 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from ..fluid.inference import Config, DataType, PlaceType, PrecisionType, Tensor, \
-    Predictor, create_predictor, get_version, get_num_bytes_of_data_type, PredictorPool
+from ..fluid.inference import Config  # noqa: F401
+from ..fluid.inference import DataType  # noqa: F401
+from ..fluid.inference import PlaceType  # noqa: F401
+from ..fluid.inference import PrecisionType  # noqa: F401
+from ..fluid.inference import Tensor  # noqa: F401
+from ..fluid.inference import Predictor  # noqa: F401
+from ..fluid.inference import create_predictor  # noqa: F401
+from ..fluid.inference import get_version  # noqa: F401
+from ..fluid.inference import get_num_bytes_of_data_type  # noqa: F401
+from ..fluid.inference import PredictorPool  # noqa: F401
+
+__all__ = [  # noqa
+    'Config',
+    'DataType',
+    'PlaceType',
+    'PrecisionType',
+    'Tensor',
+    'Predictor',
+    'create_predictor',
+    'get_version',
+    'get_num_bytes_of_data_type',
+    'PredictorPool'
+]
diff --git a/python/paddle/jit/__init__.py b/python/paddle/jit/__init__.py
index 650837b2d7702c70131250b9da94abd62b369e7a..576989e8e0d2aa019dc9ec7c7d69afa941f1dcb7 100644
--- a/python/paddle/jit/__init__.py
+++ b/python/paddle/jit/__init__.py
@@ -14,19 +14,26 @@
 
 from __future__ import print_function
 
-from ..fluid.dygraph.jit import save  #DEFINE_ALIAS
-from ..fluid.dygraph.jit import load  #DEFINE_ALIAS
-from ..fluid.dygraph.jit import TracedLayer  #DEFINE_ALIAS
-from ..fluid.dygraph.jit import set_code_level  #DEFINE_ALIAS
-from ..fluid.dygraph.jit import set_verbosity  #DEFINE_ALIAS
-from ..fluid.dygraph.jit import declarative as to_static  #DEFINE_ALIAS
-from ..fluid.dygraph.jit import not_to_static  #DEFINE_ALIAS
-from ..fluid.dygraph import ProgramTranslator  #DEFINE_ALIAS
-from ..fluid.dygraph.io import TranslatedLayer  #DEFINE_ALIAS
+from ..fluid.dygraph.jit import save  # noqa: F401
+from ..fluid.dygraph.jit import load  # noqa: F401
+from ..fluid.dygraph.jit import TracedLayer  # noqa: F401
+from ..fluid.dygraph.jit import set_code_level  # noqa: F401
+from ..fluid.dygraph.jit import set_verbosity  # noqa: F401
+from ..fluid.dygraph.jit import declarative as to_static  # noqa: F401
+from ..fluid.dygraph.jit import not_to_static  # noqa: F401
+from ..fluid.dygraph import ProgramTranslator  # noqa: F401
+from ..fluid.dygraph.io import TranslatedLayer  # noqa: F401
 
-from . import dy2static
+from . import dy2static  # noqa: F401
 
-__all__ = [
-    'save', 'load', 'TracedLayer', 'to_static', 'ProgramTranslator',
-    'TranslatedLayer', 'set_code_level', 'set_verbosity', 'not_to_static'
+__all__ = [  # noqa
+    'save',
+    'load',
+    'TracedLayer',
+    'to_static',
+    'ProgramTranslator',
+    'TranslatedLayer',
+    'set_code_level',
+    'set_verbosity',
+    'not_to_static'
 ]
diff --git a/python/paddle/jit/dy2static/__init__.py b/python/paddle/jit/dy2static/__init__.py
index 239b554180b1bd74517b152dfdf079082600b806..030d5499c2ca96d997dfe571b81c039bb0eb2c99 100644
--- a/python/paddle/jit/dy2static/__init__.py
+++ b/python/paddle/jit/dy2static/__init__.py
@@ -12,18 +12,28 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from __future__ import print_function
-
-from . import convert_operators
-from .convert_operators import *
-
-from . import convert_call_func
-from .convert_call_func import *
-
-from . import variable_trans_func
-from .variable_trans_func import *
+from .convert_call_func import convert_call  # noqa: F401
+from .convert_operators import cast_bool_if_necessary  # noqa: F401
+from .convert_operators import convert_assert  # noqa: F401
+from .convert_operators import convert_ifelse  # noqa: F401
+from .convert_operators import convert_len  # noqa: F401
+from .convert_operators import convert_logical_and  # noqa: F401
+from .convert_operators import convert_logical_not  # noqa: F401
+from .convert_operators import convert_logical_or  # noqa: F401
+from .convert_operators import convert_pop  # noqa: F401
+from .convert_operators import convert_print  # noqa: F401
+from .convert_operators import convert_shape_compare  # noqa: F401
+from .convert_operators import convert_var_dtype  # noqa: F401
+from .convert_operators import convert_var_shape  # noqa: F401
+from .convert_operators import convert_var_shape_simple  # noqa: F401
+from .convert_operators import eval_if_exist_else_none  # noqa: F401
+from .convert_operators import choose_shape_attr_or_api  # noqa: F401
+from .convert_operators import convert_while_loop  # noqa: F401
+from .variable_trans_func import create_bool_as_type  # noqa: F401
+from .variable_trans_func import create_fill_constant_node  # noqa: F401
+from .variable_trans_func import create_static_variable_gast_node  # noqa: F401
+from .variable_trans_func import data_layer_not_check  # noqa: F401
+from .variable_trans_func import to_static_variable  # noqa: F401
+from .variable_trans_func import to_static_variable_gast_node  # noqa: F401
 
 __all__ = []
-__all__ += convert_operators.__all__
-__all__ += convert_call_func.__all__
-__all__ += variable_trans_func.__all__
diff --git a/python/paddle/jit/dy2static/convert_call_func.py b/python/paddle/jit/dy2static/convert_call_func.py
index be2377608e36c75d95cb2c1c609e99cef7d438a7..4f6197a3cba6ae811998def0d59a221d2265ce0c 100644
--- a/python/paddle/jit/dy2static/convert_call_func.py
+++ b/python/paddle/jit/dy2static/convert_call_func.py
@@ -13,6 +13,6 @@
 # limitations under the License.
 from __future__ import print_function
 
-from ...fluid.dygraph.dygraph_to_static.convert_call_func import convert_call  #DEFINE_ALIAS
+from ...fluid.dygraph.dygraph_to_static.convert_call_func import convert_call  # noqa: F401
 
-__all__ = ['convert_call']
+__all__ = []
diff --git a/python/paddle/jit/dy2static/convert_operators.py b/python/paddle/jit/dy2static/convert_operators.py
index 9321cf4a0b8324cf5e312b36a17b8ab1edc72809..8d67e06d9b27a56e9aa0fc7bc57844290d1c83e1 100644
--- a/python/paddle/jit/dy2static/convert_operators.py
+++ b/python/paddle/jit/dy2static/convert_operators.py
@@ -13,27 +13,21 @@
 # limitations under the License.
 from __future__ import print_function
 
-from ...fluid.dygraph.dygraph_to_static.convert_operators import cast_bool_if_necessary  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_assert  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_ifelse  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_len  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_logical_and  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_logical_not  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_logical_or  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_pop  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_print  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_shape_compare  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_var_dtype  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_var_shape  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_var_shape_simple  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import eval_if_exist_else_none  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import choose_shape_attr_or_api  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_while_loop  #DEFINE_ALIAS
+from ...fluid.dygraph.dygraph_to_static.convert_operators import cast_bool_if_necessary  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_assert  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_ifelse  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_len  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_logical_and  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_logical_not  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_logical_or  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_pop  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_print  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_shape_compare  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_var_dtype  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_var_shape  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_var_shape_simple  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import eval_if_exist_else_none  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import choose_shape_attr_or_api  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.convert_operators import convert_while_loop  # noqa: F401
 
-__all__ = [
-    'cast_bool_if_necessary', 'convert_assert', 'convert_ifelse', 'convert_len',
-    'convert_logical_and', 'convert_logical_not', 'convert_logical_or',
-    'convert_pop', 'convert_print', 'convert_shape_compare',
-    'convert_var_dtype', 'convert_var_shape', 'convert_var_shape_simple',
-    'eval_if_exist_else_none', 'choose_shape_attr_or_api', 'convert_while_loop'
-]
+__all__ = []
diff --git a/python/paddle/jit/dy2static/variable_trans_func.py b/python/paddle/jit/dy2static/variable_trans_func.py
index 2deb1bbb0eef2542d8f8890a7fa476f370ba5e5a..9ce2bc2da381655e65225397831faa228c613ca6 100644
--- a/python/paddle/jit/dy2static/variable_trans_func.py
+++ b/python/paddle/jit/dy2static/variable_trans_func.py
@@ -14,15 +14,11 @@
 
 from __future__ import print_function
 
-from ...fluid.dygraph.dygraph_to_static.variable_trans_func import create_bool_as_type  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.variable_trans_func import create_fill_constant_node  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.variable_trans_func import create_static_variable_gast_node  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.variable_trans_func import data_layer_not_check  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.variable_trans_func import to_static_variable  #DEFINE_ALIAS
-from ...fluid.dygraph.dygraph_to_static.variable_trans_func import to_static_variable_gast_node  #DEFINE_ALIAS
+from ...fluid.dygraph.dygraph_to_static.variable_trans_func import create_bool_as_type  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.variable_trans_func import create_fill_constant_node  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.variable_trans_func import create_static_variable_gast_node  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.variable_trans_func import data_layer_not_check  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.variable_trans_func import to_static_variable  # noqa: F401
+from ...fluid.dygraph.dygraph_to_static.variable_trans_func import to_static_variable_gast_node  # noqa: F401
 
-__all__ = [
-    'create_bool_as_type', 'create_fill_constant_node',
-    'create_static_variable_gast_node', 'data_layer_not_check',
-    'to_static_variable', 'to_static_variable_gast_node'
-]
+__all__ = []
diff --git a/python/paddle/metric/__init__.py b/python/paddle/metric/__init__.py
index e41f6d76dd22159ab189654c6d30818c600b8286..2f2ef4c6f54269067406763a02e8f0772e86bc82 100644
--- a/python/paddle/metric/__init__.py
+++ b/python/paddle/metric/__init__.py
@@ -12,7 +12,18 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from .metrics import *
-from . import metrics
+from .metrics import Metric  # noqa: F401
+from .metrics import Accuracy  # noqa: F401
+from .metrics import Precision  # noqa: F401
+from .metrics import Recall  # noqa: F401
+from .metrics import Auc  # noqa: F401
+from .metrics import accuracy  # noqa: F401
 
-__all__ = metrics.__all__
+__all__ = [ #noqa
+    'Metric',
+    'Accuracy',
+    'Precision',
+    'Recall',
+    'Auc',
+    'accuracy'
+]
diff --git a/python/paddle/metric/metrics.py b/python/paddle/metric/metrics.py
index d8e400b08bd474a1a59399e413a59e9a7d77f8b7..40758fb8dc3e0f034e4d5ea9ccf6e8d2897287e1 100644
--- a/python/paddle/metric/metrics.py
+++ b/python/paddle/metric/metrics.py
@@ -26,7 +26,7 @@ from ..fluid.layers.nn import topk
 from ..fluid.framework import core, _varbase_creator, in_dygraph_mode
 import paddle
 
-__all__ = ['Metric', 'Accuracy', 'Precision', 'Recall', 'Auc', 'accuracy']
+__all__ = []
 
 
 def _is_numpy_(var):
diff --git a/python/paddle/nn/__init__.py b/python/paddle/nn/__init__.py
index 7cf3f94872de17c12910ae5453f74a18c4a1502d..3ccb9e957f4e4d6426645d7188d67159c0454a01 100644
--- a/python/paddle/nn/__init__.py
+++ b/python/paddle/nn/__init__.py
@@ -286,5 +286,6 @@ __all__ = [     #noqa
            'Swish',
            'PixelShuffle',
            'ELU',
-           'ReLU6'
+           'ReLU6',
+           'LayerDict'
 ]
diff --git a/python/paddle/nn/functional/__init__.py b/python/paddle/nn/functional/__init__.py
index d4c17a27a61780b431916b2634585de035778ce8..ff18afa9d20282a6e1147cbca0d17580456c0c95 100644
--- a/python/paddle/nn/functional/__init__.py
+++ b/python/paddle/nn/functional/__init__.py
@@ -194,5 +194,6 @@ __all__ = [     #noqa
            'embedding',
            'gather_tree',
            'one_hot',
-           'normalize'
+           'normalize',
+           'temporal_shift'
 ]
diff --git a/python/paddle/onnx/__init__.py b/python/paddle/onnx/__init__.py
index 885d1968ce1ae1ef4f6a4ff79f8ac40acb971baa..8853e78bf3d808108d540496f1d7d9e1d09121c4 100644
--- a/python/paddle/onnx/__init__.py
+++ b/python/paddle/onnx/__init__.py
@@ -12,7 +12,6 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from __future__ import print_function
-from .export import export
+from .export import export  # noqa: F401
 
 __all__ = ['export']
diff --git a/python/paddle/onnx/export.py b/python/paddle/onnx/export.py
index 4b99b42bb0423c676e8d08b1931c6488b8ab1e98..b8a217a5134fb8007f7563349c3efd40e132b0b2 100644
--- a/python/paddle/onnx/export.py
+++ b/python/paddle/onnx/export.py
@@ -15,7 +15,7 @@
 import os
 from paddle.utils import try_import
 
-__all__ = ['export']
+__all__ = []
 
 
 def export(layer, path, input_spec=None, opset_version=9, **configs):
diff --git a/python/paddle/static/__init__.py b/python/paddle/static/__init__.py
index 688bff4a678f2af224a5c541eb2331569edb717e..93394f9b5afdef5cb8a4d4f355c36e59d974fd5a 100644
--- a/python/paddle/static/__init__.py
+++ b/python/paddle/static/__init__.py
@@ -85,11 +85,21 @@ __all__ = [     #noqa
            'load',
            'save_inference_model',
            'load_inference_model',
+           'serialize_program',
+           'serialize_persistables',
+           'save_to_file',
+           'deserialize_program',
+           'deserialize_persistables',
+           'load_from_file',
            'normalize_program',
            'load_program_state',
            'set_program_state',
            'cpu_places',
            'cuda_places',
            'Variable',
-           'create_global_var'
+           'create_global_var',
+           'accuracy',
+           'auc',
+           'device_guard',
+           'create_parameter'
 ]
diff --git a/python/paddle/static/nn/__init__.py b/python/paddle/static/nn/__init__.py
index 416f6e4f3df06886dbd15c7a427b3620a1957842..b589d9f87895b73ba319499969c744f21f49c657 100644
--- a/python/paddle/static/nn/__init__.py
+++ b/python/paddle/static/nn/__init__.py
@@ -68,7 +68,6 @@ __all__ = [     #noqa
     'conv2d_transpose',
     'conv3d',
     'conv3d_transpose',
-    'create_parameter',
     'crf_decoding',
     'data_norm',
     'deform_conv2d',
diff --git a/python/paddle/tensor/__init__.py b/python/paddle/tensor/__init__.py
index c8d80fc9bc68cbbff4e270bbab4d8203e663bb2e..5aeae126d83765c56d597357edae38221389d134 100755
--- a/python/paddle/tensor/__init__.py
+++ b/python/paddle/tensor/__init__.py
@@ -25,7 +25,6 @@ from .creation import ones_like  # noqa: F401
 from .creation import zeros  # noqa: F401
 from .creation import zeros_like  # noqa: F401
 from .creation import arange  # noqa: F401
-from .creation import eye  # noqa: F401
 from .creation import full  # noqa: F401
 from .creation import full_like  # noqa: F401
 from .creation import triu  # noqa: F401
@@ -82,7 +81,6 @@ from .manipulation import squeeze  # noqa: F401
 from .manipulation import squeeze_  # noqa: F401
 from .manipulation import stack  # noqa: F401
 from .manipulation import strided_slice  # noqa: F401
-from .manipulation import transpose  # noqa: F401
 from .manipulation import unique  # noqa: F401
 from .manipulation import unsqueeze  # noqa: F401
 from .manipulation import unsqueeze_  # noqa: F401
@@ -143,7 +141,6 @@ from .math import add  # noqa: F401
 from .math import add_  # noqa: F401
 from .math import subtract  # noqa: F401
 from .math import subtract_  # noqa: F401
-from .math import atan  # noqa: F401
 from .math import logsumexp  # noqa: F401
 from .math import inverse  # noqa: F401
 from .math import log2  # noqa: F401
@@ -227,7 +224,6 @@ tensor_method_func  = [ #noqa
            'log2',
            'log10',
            'logsumexp',
-           'mul',
            'multiplex',
            'pow',
            'prod',