fix paddle.vision.models.* en (#43636)

* add outputs for code examples, test=document_fix

* missing code example update, test=document_fix

* missing returns, test=document_fix

* fix the case of proper nouns, test=document_fix

* fix SqueezeNet code example (add a blank line), test=document_fix

* fix ShuffleNetV2 args style, test=document_fix

* missing paper links, test=document_fix

* missing `optional`, test=document_fix

* refine parameter `pretrained`, test=document_fix

* add parameter `kwargs`, test=document_fix

* refine parameter `num_classes`, test=document_fix

* capitalize the first letter, test=document_fix

* missing output of code example, test=document_fix

* fix wrong api label, test=document_fix

* refine returns, test=document_fix

* add missing `.` after paper link, test=document_fix

* remove extra `:`, test=document_fix

* :math:\`0.0\`, test=document_fix

python/paddle/vision/models/alexnet.py

@@ -74,18 +74,28 @@ class ConvPoolLayer(nn.Layer):
 class AlexNet(nn.Layer):
     """AlexNet model from
     `"ImageNet Classification with Deep Convolutional Neural Networks"
-    <https://proceedings.neurips.cc/paper/2012/file/c399862d3b9d6b76c8436e924a68c45b-Paper.pdf>`_
+    <https://proceedings.neurips.cc/paper/2012/file/c399862d3b9d6b76c8436e924a68c45b-Paper.pdf>`_.
 
     Args:
-        num_classes (int): Output dim of last fc layer. Default: 1000.
+        num_classes (int, optional): Output dim of last fc layer. If num_classes <= 0, last fc layer 
+                            will not be defined. Default: 1000.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of AlexNet model.
 
     Examples:
         .. code-block:: python
 
+            import paddle
             from paddle.vision.models import AlexNet
 
             alexnet = AlexNet()
 
+            x = paddle.rand([1, 3, 224, 224])
+            out = alexnet(x)
+
+            print(out.shape)
+            # [1, 1000]
     """
 
     def __init__(self, num_classes=1000):
@@ -175,19 +185,21 @@ def _alexnet(arch, pretrained, **kwargs):
 
 
 def alexnet(pretrained=False, **kwargs):
-    """
-    AlexNet model
+    """AlexNet model from
+    `"ImageNet Classification with Deep Convolutional Neural Networks"
+    <https://proceedings.neurips.cc/paper/2012/file/c399862d3b9d6b76c8436e924a68c45b-Paper.pdf>`_.
 
     Args:
-        pretrained (bool, optional): If True, returns a model pre-trained on ImageNet. Default: False.
-        **kwargs: Additional keyword arguments，For details, please refer to :ref:`AlexNet <api_paddle_vision_models_AlexNet>`. 
-    
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`AlexNet <api_paddle_vision_AlexNet>`.
+
     Returns:
-        the model of alexnet.
+        :ref:`api_paddle_nn_Layer`. An instance of AlexNet model.
     
     Examples:
         .. code-block:: python
-          :name: code-example
+
             import paddle
             from paddle.vision.models import alexnet
 

python/paddle/vision/models/densenet.py

@@ -185,14 +185,18 @@ class ConvBNLayer(nn.Layer):
 
 class DenseNet(nn.Layer):
     """DenseNet model from
-    `"Densely Connected Convolutional Networks" <https://arxiv.org/pdf/1608.06993.pdf>`_
+    `"Densely Connected Convolutional Networks" <https://arxiv.org/pdf/1608.06993.pdf>`_.
 
     Args:
-        layers (int): layers of densenet. Default: 121.
-        bn_size (int): expansion of growth rate in the middle layer. Default: 4.
-        dropout (float): dropout rate. Default: 0..
-        num_classes (int): output dim of last fc layer. Default: 1000.
-        with_pool (bool): use pool before the last fc layer or not. Default: True.
+        layers (int, optional): Layers of DenseNet. Default: 121.
+        bn_size (int, optional): Expansion of growth rate in the middle layer. Default: 4.
+        dropout (float, optional): Dropout rate. Default: :math:`0.0`.
+        num_classes (int, optional): Output dim of last fc layer. If num_classes <= 0, last fc layer 
+                            will not be defined. Default: 1000.
+        with_pool (bool, optional): Use pool before the last fc layer or not. Default: True.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of DenseNet model.
 
     Examples:
         .. code-block:: python
@@ -207,6 +211,7 @@ class DenseNet(nn.Layer):
             out = densenet(x)
 
             print(out.shape)
+            # [1, 1000]
     """
 
     def __init__(self,
@@ -314,14 +319,21 @@ def _densenet(arch, layers, pretrained, **kwargs):
 
 
 def densenet121(pretrained=False, **kwargs):
-    """DenseNet 121-layer model
+    """DenseNet 121-layer model from
+    `"Densely Connected Convolutional Networks" <https://arxiv.org/pdf/1608.06993.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`DenseNet <api_paddle_vision_DenseNet>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of DenseNet 121-layer model.
 
     Examples:
         .. code-block:: python
 
+            import paddle
             from paddle.vision.models import densenet121
 
             # build model
@@ -329,15 +341,27 @@ def densenet121(pretrained=False, **kwargs):
 
             # build model and load imagenet pretrained weight
             # model = densenet121(pretrained=True)
+
+            x = paddle.rand([1, 3, 224, 224])
+            out = model(x)
+
+            print(out.shape)
+            # [1, 1000]
     """
     return _densenet('densenet121', 121, pretrained, **kwargs)
 
 
 def densenet161(pretrained=False, **kwargs):
-    """DenseNet 161-layer model
+    """DenseNet 161-layer model from
+    `"Densely Connected Convolutional Networks" <https://arxiv.org/pdf/1608.06993.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`DenseNet <api_paddle_vision_DenseNet>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of DenseNet 161-layer model.
 
     Examples:
         .. code-block:: python
@@ -354,14 +378,21 @@ def densenet161(pretrained=False, **kwargs):
 
 
 def densenet169(pretrained=False, **kwargs):
-    """DenseNet 169-layer model
+    """DenseNet 169-layer model from
+    `"Densely Connected Convolutional Networks" <https://arxiv.org/pdf/1608.06993.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`DenseNet <api_paddle_vision_DenseNet>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of DenseNet 169-layer model.
 
     Examples:
         .. code-block:: python
 
+            import paddle
             from paddle.vision.models import densenet169
 
             # build model
@@ -369,19 +400,32 @@ def densenet169(pretrained=False, **kwargs):
 
             # build model and load imagenet pretrained weight
             # model = densenet169(pretrained=True)
+
+            x = paddle.rand([1, 3, 224, 224])
+            out = model(x)
+
+            print(out.shape)
+            # [1, 1000]
     """
     return _densenet('densenet169', 169, pretrained, **kwargs)
 
 
 def densenet201(pretrained=False, **kwargs):
-    """DenseNet 201-layer model
+    """DenseNet 201-layer model from
+    `"Densely Connected Convolutional Networks" <https://arxiv.org/pdf/1608.06993.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`DenseNet <api_paddle_vision_DenseNet>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of DenseNet 201-layer model.
 
     Examples:
         .. code-block:: python
 
+            import paddle
             from paddle.vision.models import densenet201
 
             # build model
@@ -389,19 +433,31 @@ def densenet201(pretrained=False, **kwargs):
 
             # build model and load imagenet pretrained weight
             # model = densenet201(pretrained=True)
+            x = paddle.rand([1, 3, 224, 224])
+            out = model(x)
+
+            print(out.shape)
+            # [1, 1000]
     """
     return _densenet('densenet201', 201, pretrained, **kwargs)
 
 
 def densenet264(pretrained=False, **kwargs):
-    """DenseNet 264-layer model
+    """DenseNet 264-layer model from
+    `"Densely Connected Convolutional Networks" <https://arxiv.org/pdf/1608.06993.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`DenseNet <api_paddle_vision_DenseNet>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of DenseNet 264-layer model.
 
     Examples:
         .. code-block:: python
 
+            import paddle
             from paddle.vision.models import densenet264
 
             # build model
@@ -409,5 +465,11 @@ def densenet264(pretrained=False, **kwargs):
 
             # build model and load imagenet pretrained weight
             # model = densenet264(pretrained=True)
+
+            x = paddle.rand([1, 3, 224, 224])
+            out = model(x)
+
+            print(out.shape)
+            # [1, 1000]
     """
     return _densenet('densenet264', 264, pretrained, **kwargs)

python/paddle/vision/models/googlenet.py

@@ -97,12 +97,15 @@ class Inception(nn.Layer):
 
 class GoogLeNet(nn.Layer):
     """GoogLeNet (Inception v1) model architecture from
-    `"Going Deeper with Convolutions" <https://arxiv.org/pdf/1409.4842.pdf>`_
+    `"Going Deeper with Convolutions" <https://arxiv.org/pdf/1409.4842.pdf>`_.
     
     Args:
-        num_classes (int): output dim of last fc layer. If num_classes <=0, last fc layer 
+        num_classes (int, optional): Output dim of last fc layer. If num_classes <= 0, last fc layer 
                             will not be defined. Default: 1000.
-        with_pool (bool, optional): use pool before the last fc layer or not. Default: True.
+        with_pool (bool, optional): Use pool before the last fc layer or not. Default: True.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of GoogLeNet (Inception v1) model.
 
     Examples:
         .. code-block:: python
@@ -116,7 +119,8 @@ class GoogLeNet(nn.Layer):
             x = paddle.rand([1, 3, 224, 224])
             out, out1, out2 = model(x)
 
-            print(out.shape)
+            print(out.shape, out1.shape, out2.shape)
+            # [1, 1000] [1, 1000] [1, 1000]
     """
 
     def __init__(self, num_classes=1000, with_pool=True):
@@ -219,10 +223,15 @@ class GoogLeNet(nn.Layer):
 
 def googlenet(pretrained=False, **kwargs):
     """GoogLeNet (Inception v1) model architecture from
-    `"Going Deeper with Convolutions" <https://arxiv.org/pdf/1409.4842.pdf>`_
+    `"Going Deeper with Convolutions" <https://arxiv.org/pdf/1409.4842.pdf>`_.
     
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`GoogLeNet <api_paddle_vision_GoogLeNet>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of GoogLeNet (Inception v1) model.
 
     Examples:
         .. code-block:: python
@@ -239,7 +248,8 @@ def googlenet(pretrained=False, **kwargs):
             x = paddle.rand([1, 3, 224, 224])
             out, out1, out2 = model(x)
 
-            print(out.shape)
+            print(out.shape, out1.shape, out2.shape)
+            # [1, 1000] [1, 1000] [1, 1000]
     """
     model = GoogLeNet(**kwargs)
     arch = "googlenet"

python/paddle/vision/models/inceptionv3.py

@@ -413,12 +413,16 @@ class InceptionE(nn.Layer):
 
 
 class InceptionV3(nn.Layer):
-    """
-    InceptionV3
+    """Inception v3 model from
+    `"Rethinking the Inception Architecture for Computer Vision" <https://arxiv.org/pdf/1512.00567.pdf>`_.
+
     Args:
-        num_classes (int, optional): output dim of last fc layer. If num_classes <=0, last fc layer 
-                            will not be defined. Default: 1000. 
-        with_pool (bool, optional): use pool before the last fc layer or not. Default: True.
+        num_classes (int, optional): Output dim of last fc layer. If num_classes <= 0, last fc layer 
+                            will not be defined. Default: 1000.
+        with_pool (bool, optional): Use pool before the last fc layer or not. Default: True.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of Inception v3 model.
 
     Examples:
         .. code-block:: python
@@ -432,6 +436,7 @@ class InceptionV3(nn.Layer):
             out = inception_v3(x)
 
             print(out.shape)
+            # [1, 1000]
     """
 
     def __init__(self, num_classes=1000, with_pool=True):
@@ -505,13 +510,17 @@ class InceptionV3(nn.Layer):
 
 
 def inception_v3(pretrained=False, **kwargs):
-    """
-    InceptionV3 model from
-    `"Rethinking the Inception Architecture for Computer Vision" <https://arxiv.org/pdf/1512.00567.pdf>`_
+    """Inception v3 model from
+    `"Rethinking the Inception Architecture for Computer Vision" <https://arxiv.org/pdf/1512.00567.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet
-    
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`InceptionV3 <api_paddle_vision_InceptionV3>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of Inception v3 model.
+
     Examples:
         .. code-block:: python
 
@@ -528,6 +537,7 @@ def inception_v3(pretrained=False, **kwargs):
             out = model(x)
 
             print(out.shape)
+            # [1, 1000]
     """
     model = InceptionV3(**kwargs)
     arch = "inception_v3"

python/paddle/vision/models/lenet.py

@@ -20,18 +20,28 @@ __all__ = []
 
 class LeNet(nn.Layer):
     """LeNet model from
-    `"LeCun Y, Bottou L, Bengio Y, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324.`_
+    `"Gradient-based learning applied to document recognition" <https://ieeexplore.ieee.org/document/726791>`_.
 
     Args:
-        num_classes (int): output dim of last fc layer. If num_classes <=0, last fc layer 
+        num_classes (int, optional): Output dim of last fc layer. If num_classes <= 0, last fc layer 
                             will not be defined. Default: 10.
 
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of LeNet model.
+
     Examples:
         .. code-block:: python
 
+            import paddle
             from paddle.vision.models import LeNet
 
             model = LeNet()
+
+            x = paddle.rand([1, 1, 28, 28])
+            out = model(x)
+
+            print(out.shape)
+            # [1, 10]
     """
 
     def __init__(self, num_classes=10):

python/paddle/vision/models/mobilenetv1.py

@@ -58,14 +58,17 @@ class MobileNetV1(nn.Layer):
     `"MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications" <https://arxiv.org/abs/1704.04861>`_.
 
     Args:
-        scale (float, optional): scale of channels in each layer. Default: 1.0.
-        num_classes (int, optional): output dim of last fc layer. If num_classes <=0, last fc layer 
+        scale (float, optional): Scale of channels in each layer. Default: 1.0.
+        num_classes (int, optional): Output dim of last fc layer. If num_classes <= 0, last fc layer 
                             will not be defined. Default: 1000.
-        with_pool (bool, optional): use pool before the last fc layer or not. Default: True.
+        with_pool (bool, optional): Use pool before the last fc layer or not. Default: True.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of MobileNetV1 model.
 
     Examples:
         .. code-block:: python
-          :name: code-example1
+
             import paddle
             from paddle.vision.models import MobileNetV1
 
@@ -217,11 +220,17 @@ def _mobilenet(arch, pretrained=False, **kwargs):
 
 
 def mobilenet_v1(pretrained=False, scale=1.0, **kwargs):
-    """MobileNetV1
+    """MobileNetV1 from
+    `"MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications" <https://arxiv.org/abs/1704.04861>`_.
     
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
-        scale: (float): scale of channels in each layer. Default: 1.0.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        scale (float, optional): Scale of channels in each layer. Default: 1.0.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`MobileNetV1 <api_paddle_vision_MobileNetV1>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of MobileNetV1 model.
 
     Examples:
         .. code-block:: python
@@ -242,6 +251,7 @@ def mobilenet_v1(pretrained=False, scale=1.0, **kwargs):
             out = model(x)
 
             print(out.shape)
+            # [1, 1000]
     """
     model = _mobilenet('mobilenetv1_' + str(scale),
                        pretrained,

python/paddle/vision/models/mobilenetv2.py

@@ -75,14 +75,17 @@ class MobileNetV2(nn.Layer):
     `"MobileNetV2: Inverted Residuals and Linear Bottlenecks" <https://arxiv.org/abs/1801.04381>`_.
 
     Args:
-        scale (float, optional): scale of channels in each layer. Default: 1.0.
-        num_classes (int, optional): output dim of last fc layer. If num_classes <=0, last fc layer 
+        scale (float, optional): Scale of channels in each layer. Default: 1.0.
+        num_classes (int, optional): Output dim of last fc layer. If num_classes <= 0, last fc layer 
                             will not be defined. Default: 1000.
-        with_pool (bool, optional): use pool before the last fc layer or not. Default: True.
+        with_pool (bool, optional): Use pool before the last fc layer or not. Default: True.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of MobileNetV2 model.
 
     Examples:
         .. code-block:: python
-          :name: code-example1
+
             import paddle
             from paddle.vision.models import MobileNetV2
 
@@ -181,11 +184,17 @@ def _mobilenet(arch, pretrained=False, **kwargs):
 
 
 def mobilenet_v2(pretrained=False, scale=1.0, **kwargs):
-    """MobileNetV2
+    """MobileNetV2 from
+    `"MobileNetV2: Inverted Residuals and Linear Bottlenecks" <https://arxiv.org/abs/1801.04381>`_.
     
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
-        scale: (float): scale of channels in each layer. Default: 1.0.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        scale (float, optional): Scale of channels in each layer. Default: 1.0.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`MobileNetV2 <api_paddle_vision_MobileNetV2>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of MobileNetV2 model.
 
     Examples:
         .. code-block:: python
@@ -206,6 +215,7 @@ def mobilenet_v2(pretrained=False, scale=1.0, **kwargs):
             out = model(x)
 
             print(out.shape)
+            # [1, 1000]
     """
     model = _mobilenet('mobilenetv2_' + str(scale),
                        pretrained,

python/paddle/vision/models/mobilenetv3.py

@@ -254,10 +254,13 @@ class MobileNetV3Small(MobileNetV3):
 
     Args:
         scale (float, optional): Scale of channels in each layer. Default: 1.0.
-        num_classes (int, optional): Output dim of last fc layer. If num_classes <=0, last fc layer
+        num_classes (int, optional): Output dim of last fc layer. If num_classes <= 0, last fc layer 
                             will not be defined. Default: 1000.
         with_pool (bool, optional): Use pool before the last fc layer or not. Default: True.
 
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of MobileNetV3 Small architecture model.
+
     Examples:
         .. code-block:: python
 
@@ -271,6 +274,7 @@ class MobileNetV3Small(MobileNetV3):
             out = model(x)
 
             print(out.shape)
+            # [1, 1000]
     """
 
     def __init__(self, scale=1.0, num_classes=1000, with_pool=True):
@@ -301,10 +305,13 @@ class MobileNetV3Large(MobileNetV3):
 
     Args:
         scale (float, optional): Scale of channels in each layer. Default: 1.0.
-        num_classes (int, optional): Output dim of last fc layer. If num_classes <=0, last fc layer
+        num_classes (int, optional): Output dim of last fc layer. If num_classes <= 0, last fc layer 
                             will not be defined. Default: 1000.
         with_pool (bool, optional): Use pool before the last fc layer or not. Default: True.
 
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of MobileNetV3 Large architecture model.
+
     Examples:
         .. code-block:: python
 
@@ -318,6 +325,7 @@ class MobileNetV3Large(MobileNetV3):
             out = model(x)
 
             print(out.shape)
+            # [1, 1000]
     """
 
     def __init__(self, scale=1.0, num_classes=1000, with_pool=True):
@@ -379,8 +387,13 @@ def mobilenet_v3_small(pretrained=False, scale=1.0, **kwargs):
     `"Searching for MobileNetV3" <https://arxiv.org/abs/1905.02244>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
         scale (float, optional): Scale of channels in each layer. Default: 1.0.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`MobileNetV3Small <api_paddle_vision_MobileNetV3Small>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of MobileNetV3 Small architecture model.
 
     Examples:
         .. code-block:: python
@@ -401,7 +414,7 @@ def mobilenet_v3_small(pretrained=False, scale=1.0, **kwargs):
             out = model(x)
 
             print(out.shape)
-
+            # [1, 1000]
     """
     model = _mobilenet_v3("mobilenet_v3_small",
                           scale=scale,
@@ -415,8 +428,13 @@ def mobilenet_v3_large(pretrained=False, scale=1.0, **kwargs):
     `"Searching for MobileNetV3" <https://arxiv.org/abs/1905.02244>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
         scale (float, optional): Scale of channels in each layer. Default: 1.0.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`MobileNetV3Large <api_paddle_vision_MobileNetV3Large>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of MobileNetV3 Large architecture model.
 
     Examples:
         .. code-block:: python
@@ -437,7 +455,7 @@ def mobilenet_v3_large(pretrained=False, scale=1.0, **kwargs):
             out = model(x)
 
             print(out.shape)
-
+            # [1, 1000]
     """
     model = _mobilenet_v3("mobilenet_v3_large",
                           scale=scale,

python/paddle/vision/models/resnet.py

@@ -177,19 +177,19 @@ class BottleneckBlock(nn.Layer):
 
 class ResNet(nn.Layer):
     """ResNet model from
-    `"Deep Residual Learning for Image Recognition" <https://arxiv.org/pdf/1512.03385.pdf>`_
+    `"Deep Residual Learning for Image Recognition" <https://arxiv.org/pdf/1512.03385.pdf>`_.
 
     Args:
-        Block (BasicBlock|BottleneckBlock): block module of model.
-        depth (int, optional): layers of ResNet, Default: 50.
-        width (int, optional): base width per convolution group for each convolution block, Default: 64.
-        num_classes (int, optional): output dim of last fc layer. If num_classes <=0, last fc layer
+        Block (BasicBlock|BottleneckBlock): Block module of model.
+        depth (int, optional): Layers of ResNet, Default: 50.
+        width (int, optional): Base width per convolution group for each convolution block, Default: 64.
+        num_classes (int, optional): Output dim of last fc layer. If num_classes <= 0, last fc layer 
                             will not be defined. Default: 1000.
-        with_pool (bool, optional): use pool before the last fc layer or not. Default: True.
-        groups (int, optional): number of groups for each convolution block, Default: 1.
+        with_pool (bool, optional): Use pool before the last fc layer or not. Default: True.
+        groups (int, optional): Number of groups for each convolution block, Default: 1.
 
     Returns:
-        ResNet model. An instance of :ref:`api_fluid_dygraph_Layer`.
+        :ref:`api_paddle_nn_Layer`. An instance of ResNet model.
 
     Examples:
         .. code-block:: python
@@ -215,7 +215,6 @@ class ResNet(nn.Layer):
 
             print(out.shape)
             # [1, 1000]
-
     """
 
     def __init__(self,
@@ -330,14 +329,15 @@ def _resnet(arch, Block, depth, pretrained, **kwargs):
 
 def resnet18(pretrained=False, **kwargs):
     """ResNet 18-layer model from
-    `"Deep Residual Learning for Image Recognition" <https://arxiv.org/pdf/1512.03385.pdf>`_
+    `"Deep Residual Learning for Image Recognition" <https://arxiv.org/pdf/1512.03385.pdf>`_.
 
     Args:
         pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
                             on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ResNet <api_paddle_vision_ResNet>`.
 
     Returns:
-        ResNet 18-layer model. An instance of :ref:`api_fluid_dygraph_Layer`.
+        :ref:`api_paddle_nn_Layer`. An instance of ResNet 18-layer model.
 
     Examples:
         .. code-block:: python
@@ -362,14 +362,15 @@ def resnet18(pretrained=False, **kwargs):
 
 def resnet34(pretrained=False, **kwargs):
     """ResNet 34-layer model from
-    `"Deep Residual Learning for Image Recognition" <https://arxiv.org/pdf/1512.03385.pdf>`_
+    `"Deep Residual Learning for Image Recognition" <https://arxiv.org/pdf/1512.03385.pdf>`_.
 
     Args:
         pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
                             on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ResNet <api_paddle_vision_ResNet>`.
 
     Returns:
-        ResNet 34-layer model. An instance of :ref:`api_fluid_dygraph_Layer`.
+        :ref:`api_paddle_nn_Layer`. An instance of ResNet 34-layer model.
 
     Examples:
         .. code-block:: python
@@ -394,14 +395,15 @@ def resnet34(pretrained=False, **kwargs):
 
 def resnet50(pretrained=False, **kwargs):
     """ResNet 50-layer model from
-    `"Deep Residual Learning for Image Recognition" <https://arxiv.org/pdf/1512.03385.pdf>`_
+    `"Deep Residual Learning for Image Recognition" <https://arxiv.org/pdf/1512.03385.pdf>`_.
 
     Args:
         pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
                             on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ResNet <api_paddle_vision_ResNet>`.
 
     Returns:
-        ResNet 50-layer model. An instance of :ref:`api_fluid_dygraph_Layer`.
+        :ref:`api_paddle_nn_Layer`. An instance of ResNet 50-layer model.
 
     Examples:
         .. code-block:: python
@@ -426,14 +428,15 @@ def resnet50(pretrained=False, **kwargs):
 
 def resnet101(pretrained=False, **kwargs):
     """ResNet 101-layer model from
-    `"Deep Residual Learning for Image Recognition" <https://arxiv.org/pdf/1512.03385.pdf>`_
+    `"Deep Residual Learning for Image Recognition" <https://arxiv.org/pdf/1512.03385.pdf>`_.
 
     Args:
         pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
                             on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ResNet <api_paddle_vision_ResNet>`.
 
     Returns:
-        ResNet 101-layer. An instance of :ref:`api_fluid_dygraph_Layer`.
+        :ref:`api_paddle_nn_Layer`. An instance of ResNet 101-layer.
 
     Examples:
         .. code-block:: python
@@ -458,14 +461,15 @@ def resnet101(pretrained=False, **kwargs):
 
 def resnet152(pretrained=False, **kwargs):
     """ResNet 152-layer model from
-    `"Deep Residual Learning for Image Recognition" <https://arxiv.org/pdf/1512.03385.pdf>`_
+    `"Deep Residual Learning for Image Recognition" <https://arxiv.org/pdf/1512.03385.pdf>`_.
 
     Args:
         pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
                             on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ResNet <api_paddle_vision_ResNet>`.
 
     Returns:
-        ResNet 152-layer model. An instance of :ref:`api_fluid_dygraph_Layer`.
+        :ref:`api_paddle_nn_Layer`. An instance of ResNet 152-layer model.
 
     Examples:
         .. code-block:: python
@@ -490,14 +494,15 @@ def resnet152(pretrained=False, **kwargs):
 
 def resnext50_32x4d(pretrained=False, **kwargs):
     """ResNeXt-50 32x4d model from
-    `"Aggregated Residual Transformations for Deep Neural Networks" <https://arxiv.org/pdf/1611.05431.pdf>`_
+    `"Aggregated Residual Transformations for Deep Neural Networks" <https://arxiv.org/pdf/1611.05431.pdf>`_.
     
     Args:
         pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
                             on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ResNet <api_paddle_vision_ResNet>`.
 
     Returns:
-        ResNeXt-50 32x4d model. An instance of :ref:`api_fluid_dygraph_Layer`.
+        :ref:`api_paddle_nn_Layer`. An instance of ResNeXt-50 32x4d model.
 
     Examples:
         .. code-block:: python
@@ -524,14 +529,15 @@ def resnext50_32x4d(pretrained=False, **kwargs):
 
 def resnext50_64x4d(pretrained=False, **kwargs):
     """ResNeXt-50 64x4d model from
-    `"Aggregated Residual Transformations for Deep Neural Networks" <https://arxiv.org/pdf/1611.05431.pdf>`_
+    `"Aggregated Residual Transformations for Deep Neural Networks" <https://arxiv.org/pdf/1611.05431.pdf>`_.
     
     Args:
         pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
                             on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ResNet <api_paddle_vision_ResNet>`.
 
     Returns:
-        ResNeXt-50 64x4d model. An instance of :ref:`api_fluid_dygraph_Layer`.
+        :ref:`api_paddle_nn_Layer`. An instance of ResNeXt-50 64x4d model.
 
     Examples:
         .. code-block:: python
@@ -558,14 +564,15 @@ def resnext50_64x4d(pretrained=False, **kwargs):
 
 def resnext101_32x4d(pretrained=False, **kwargs):
     """ResNeXt-101 32x4d model from
-    `"Aggregated Residual Transformations for Deep Neural Networks" <https://arxiv.org/pdf/1611.05431.pdf>`_
+    `"Aggregated Residual Transformations for Deep Neural Networks" <https://arxiv.org/pdf/1611.05431.pdf>`_.
     
     Args:
         pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
                             on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ResNet <api_paddle_vision_ResNet>`.
 
     Returns:
-        ResNeXt-101 32x4d model. An instance of :ref:`api_fluid_dygraph_Layer`.
+        :ref:`api_paddle_nn_Layer`. An instance of ResNeXt-101 32x4d model.
 
     Examples:
         .. code-block:: python
@@ -593,14 +600,15 @@ def resnext101_32x4d(pretrained=False, **kwargs):
 
 def resnext101_64x4d(pretrained=False, **kwargs):
     """ResNeXt-101 64x4d model from
-    `"Aggregated Residual Transformations for Deep Neural Networks" <https://arxiv.org/pdf/1611.05431.pdf>`_
+    `"Aggregated Residual Transformations for Deep Neural Networks" <https://arxiv.org/pdf/1611.05431.pdf>`_.
     
     Args:
         pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
                             on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ResNet <api_paddle_vision_ResNet>`.
 
     Returns:
-        ResNeXt-101 64x4d model. An instance of :ref:`api_fluid_dygraph_Layer`.
+        :ref:`api_paddle_nn_Layer`. An instance of ResNeXt-101 64x4d model.
 
     Examples:
         .. code-block:: python
@@ -628,14 +636,15 @@ def resnext101_64x4d(pretrained=False, **kwargs):
 
 def resnext152_32x4d(pretrained=False, **kwargs):
     """ResNeXt-152 32x4d model from
-    `"Aggregated Residual Transformations for Deep Neural Networks" <https://arxiv.org/pdf/1611.05431.pdf>`_
+    `"Aggregated Residual Transformations for Deep Neural Networks" <https://arxiv.org/pdf/1611.05431.pdf>`_.
     
     Args:
         pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
                             on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ResNet <api_paddle_vision_ResNet>`.
 
     Returns:
-        ResNeXt-152 32x4d model. An instance of :ref:`api_fluid_dygraph_Layer`.
+        :ref:`api_paddle_nn_Layer`. An instance of ResNeXt-152 32x4d model.
 
     Examples:
         .. code-block:: python
@@ -663,14 +672,15 @@ def resnext152_32x4d(pretrained=False, **kwargs):
 
 def resnext152_64x4d(pretrained=False, **kwargs):
     """ResNeXt-152 64x4d model from
-    `"Aggregated Residual Transformations for Deep Neural Networks" <https://arxiv.org/pdf/1611.05431.pdf>`_
+    `"Aggregated Residual Transformations for Deep Neural Networks" <https://arxiv.org/pdf/1611.05431.pdf>`_.
     
     Args:
         pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
                             on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ResNet <api_paddle_vision_ResNet>`.
 
     Returns:
-        ResNeXt-152 64x4d model. An instance of :ref:`api_fluid_dygraph_Layer`.
+        :ref:`api_paddle_nn_Layer`. An instance of ResNeXt-152 64x4d model.
 
     Examples:
         .. code-block:: python
@@ -703,9 +713,10 @@ def wide_resnet50_2(pretrained=False, **kwargs):
     Args:
         pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
                             on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ResNet <api_paddle_vision_ResNet>`.
 
     Returns:
-        Wide ResNet-50-2 model. An instance of :ref:`api_fluid_dygraph_Layer`.
+        :ref:`api_paddle_nn_Layer`. An instance of Wide ResNet-50-2 model.
 
     Examples:
         .. code-block:: python
@@ -736,9 +747,10 @@ def wide_resnet101_2(pretrained=False, **kwargs):
     Args:
         pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
                             on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ResNet <api_paddle_vision_ResNet>`.
 
     Returns:
-        Wide ResNet-101-2 model. An instance of :ref:`api_fluid_dygraph_Layer`.
+        :ref:`api_paddle_nn_Layer`. An instance of Wide ResNet-101-2 model.
 
     Examples:
         .. code-block:: python

python/paddle/vision/models/shufflenetv2.py

@@ -190,14 +190,17 @@ class InvertedResidualDS(nn.Layer):
 
 class ShuffleNetV2(nn.Layer):
     """ShuffleNetV2 model from
-    `"ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_
+    `"ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_.
 
     Args:
-        scale (float, optional) - scale of output channels. Default: True.
-        act (str, optional) - activation function of neural network. Default: "relu".
-        num_classes (int, optional): output dim of last fc layer. If num_classes <=0, last fc layer
+        scale (float, optional): Scale of output channels. Default: True.
+        act (str, optional): Activation function of neural network. Default: "relu".
+        num_classes (int, optional): Output dim of last fc layer. If num_classes <= 0, last fc layer 
                             will not be defined. Default: 1000.
-        with_pool (bool, optional): use pool before the last fc layer or not. Default: True.
+        with_pool (bool, optional): Use pool before the last fc layer or not. Default: True.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of ShuffleNetV2 model.
 
     Examples:
         .. code-block:: python
@@ -209,7 +212,7 @@ class ShuffleNetV2(nn.Layer):
             x = paddle.rand([1, 3, 224, 224])
             out = shufflenet_v2_swish(x)
             print(out.shape)
-
+            # [1, 1000]
     """
 
     def __init__(self, scale=1.0, act="relu", num_classes=1000, with_pool=True):
@@ -315,10 +318,15 @@ def _shufflenet_v2(arch, pretrained=False, **kwargs):
 
 def shufflenet_v2_x0_25(pretrained=False, **kwargs):
     """ShuffleNetV2 with 0.25x output channels, as described in
-    `"ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_ 。
+    `"ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ShuffleNetV2 <api_paddle_vision_ShuffleNetV2>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of ShuffleNetV2 with 0.25x output channels.
 
     Examples:
         .. code-block:: python
@@ -336,7 +344,7 @@ def shufflenet_v2_x0_25(pretrained=False, **kwargs):
             out = model(x)
 
             print(out.shape)
-
+            # [1, 1000]
     """
     return _shufflenet_v2("shufflenet_v2_x0_25",
                           scale=0.25,
@@ -346,10 +354,15 @@ def shufflenet_v2_x0_25(pretrained=False, **kwargs):
 
 def shufflenet_v2_x0_33(pretrained=False, **kwargs):
     """ShuffleNetV2 with 0.33x output channels, as described in
-    `"ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_ 。
+    `"ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ShuffleNetV2 <api_paddle_vision_ShuffleNetV2>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of ShuffleNetV2 with 0.33x output channels.
 
     Examples:
         .. code-block:: python
@@ -367,7 +380,7 @@ def shufflenet_v2_x0_33(pretrained=False, **kwargs):
             out = model(x)
 
             print(out.shape)
-
+            # [1, 1000]
     """
     return _shufflenet_v2("shufflenet_v2_x0_33",
                           scale=0.33,
@@ -377,10 +390,15 @@ def shufflenet_v2_x0_33(pretrained=False, **kwargs):
 
 def shufflenet_v2_x0_5(pretrained=False, **kwargs):
     """ShuffleNetV2 with 0.5x output channels, as described in
-    `"ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_ 。
+    `"ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ShuffleNetV2 <api_paddle_vision_ShuffleNetV2>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of ShuffleNetV2 with 0.5x output channels.
 
     Examples:
         .. code-block:: python
@@ -398,7 +416,7 @@ def shufflenet_v2_x0_5(pretrained=False, **kwargs):
             out = model(x)
 
             print(out.shape)
-
+            # [1, 1000]
     """
     return _shufflenet_v2("shufflenet_v2_x0_5",
                           scale=0.5,
@@ -408,10 +426,15 @@ def shufflenet_v2_x0_5(pretrained=False, **kwargs):
 
 def shufflenet_v2_x1_0(pretrained=False, **kwargs):
     """ShuffleNetV2 with 1.0x output channels, as described in
-    `"ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_ 。
+    `"ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ShuffleNetV2 <api_paddle_vision_ShuffleNetV2>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of ShuffleNetV2 with 1.0x output channels.
 
     Examples:
         .. code-block:: python
@@ -429,7 +452,7 @@ def shufflenet_v2_x1_0(pretrained=False, **kwargs):
             out = model(x)
 
             print(out.shape)
-
+            # [1, 1000]
     """
     return _shufflenet_v2("shufflenet_v2_x1_0",
                           scale=1.0,
@@ -439,10 +462,15 @@ def shufflenet_v2_x1_0(pretrained=False, **kwargs):
 
 def shufflenet_v2_x1_5(pretrained=False, **kwargs):
     """ShuffleNetV2 with 1.5x output channels, as described in
-    `"ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_ 。
+    `"ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ShuffleNetV2 <api_paddle_vision_ShuffleNetV2>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of ShuffleNetV2 with 1.5x output channels.
 
     Examples:
         .. code-block:: python
@@ -460,7 +488,7 @@ def shufflenet_v2_x1_5(pretrained=False, **kwargs):
             out = model(x)
 
             print(out.shape)
-
+            # [1, 1000]
     """
     return _shufflenet_v2("shufflenet_v2_x1_5",
                           scale=1.5,
@@ -470,10 +498,15 @@ def shufflenet_v2_x1_5(pretrained=False, **kwargs):
 
 def shufflenet_v2_x2_0(pretrained=False, **kwargs):
     """ShuffleNetV2 with 2.0x output channels, as described in
-    `"ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_ 。
+    `"ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ShuffleNetV2 <api_paddle_vision_ShuffleNetV2>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of ShuffleNetV2 with 2.0x output channels.
 
     Examples:
         .. code-block:: python
@@ -491,7 +524,7 @@ def shufflenet_v2_x2_0(pretrained=False, **kwargs):
             out = model(x)
 
             print(out.shape)
-
+            # [1, 1000]
     """
     return _shufflenet_v2("shufflenet_v2_x2_0",
                           scale=2.0,
@@ -500,11 +533,16 @@ def shufflenet_v2_x2_0(pretrained=False, **kwargs):
 
 
 def shufflenet_v2_swish(pretrained=False, **kwargs):
-    """ShuffleNetV2 with 1.0x output channels and swish activation function, as described in
-    `"ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_ 。
+    """ShuffleNetV2 with swish activation function, as described in
+    `"ShuffleNet V2: Practical Guidelines for Ecient CNN Architecture Design" <https://arxiv.org/pdf/1807.11164.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`ShuffleNetV2 <api_paddle_vision_ShuffleNetV2>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of ShuffleNetV2 with swish activation function.
 
     Examples:
         .. code-block:: python
@@ -522,7 +560,7 @@ def shufflenet_v2_swish(pretrained=False, **kwargs):
             out = model(x)
 
             print(out.shape)
-
+            # [1, 1000]
     """
     return _shufflenet_v2("shufflenet_v2_swish",
                           scale=1.0,

python/paddle/vision/models/squeezenet.py

@@ -76,15 +76,20 @@ class MakeFire(nn.Layer):
 class SqueezeNet(nn.Layer):
     """SqueezeNet model from
     `"SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size"
-    <https://arxiv.org/pdf/1602.07360.pdf>`_
+    <https://arxiv.org/pdf/1602.07360.pdf>`_.
 
     Args:
-        version (str): version of squeezenet, which can be "1.0" or "1.1".
-        num_classes (int): output dim of last fc layer. Default: 1000.
-        with_pool (bool): use pool before the last fc layer or not. Default: True.
+        version (str): Version of SqueezeNet, which can be "1.0" or "1.1".
+        num_classes (int, optional): Output dim of last fc layer. If num_classes <= 0, last fc layer 
+                            will not be defined. Default: 1000.
+        with_pool (bool, optional): Use pool before the last fc layer or not. Default: True.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of SqueezeNet model.
 
     Examples:
         .. code-block:: python
+
             import paddle
             from paddle.vision.models import SqueezeNet
 
@@ -98,7 +103,7 @@ class SqueezeNet(nn.Layer):
             out = model(x)
 
             print(out.shape)
-
+            # [1, 1000]
     """
 
     def __init__(self, version, num_classes=1000, with_pool=True):
@@ -205,14 +210,22 @@ def _squeezenet(arch, version, pretrained, **kwargs):
 
 
 def squeezenet1_0(pretrained=False, **kwargs):
-    """SqueezeNet v1.0 model
+    """SqueezeNet v1.0 model from
+    `"SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size"
+    <https://arxiv.org/pdf/1602.07360.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`SqueezeNet <api_paddle_vision_SqueezeNet>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of SqueezeNet v1.0 model.
 
     Examples:
         .. code-block:: python
 
+            import paddle
             from paddle.vision.models import squeezenet1_0
 
             # build model
@@ -220,19 +233,33 @@ def squeezenet1_0(pretrained=False, **kwargs):
 
             # build model and load imagenet pretrained weight
             # model = squeezenet1_0(pretrained=True)
+
+            x = paddle.rand([1, 3, 224, 224])
+            out = model(x)
+
+            print(out.shape)
+            # [1, 1000]
     """
     return _squeezenet('squeezenet1_0', '1.0', pretrained, **kwargs)
 
 
 def squeezenet1_1(pretrained=False, **kwargs):
-    """SqueezeNet v1.1 model
+    """SqueezeNet v1.1 model from
+    `"SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size"
+    <https://arxiv.org/pdf/1602.07360.pdf>`_.
 
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`SqueezeNet <api_paddle_vision_SqueezeNet>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of SqueezeNet v1.1 model.
 
     Examples:
         .. code-block:: python
 
+            import paddle
             from paddle.vision.models import squeezenet1_1
 
             # build model
@@ -240,5 +267,11 @@ def squeezenet1_1(pretrained=False, **kwargs):
 
             # build model and load imagenet pretrained weight
             # model = squeezenet1_1(pretrained=True)
+
+            x = paddle.rand([1, 3, 224, 224])
+            out = model(x)
+
+            print(out.shape)
+            # [1, 1000]
     """
     return _squeezenet('squeezenet1_1', '1.1', pretrained, **kwargs)

python/paddle/vision/models/vgg.py

@@ -29,17 +29,19 @@ model_urls = {
 
 class VGG(nn.Layer):
     """VGG model from
-    `"Very Deep Convolutional Networks For Large-Scale Image Recognition" <https://arxiv.org/pdf/1409.1556.pdf>`_
+    `"Very Deep Convolutional Networks For Large-Scale Image Recognition" <https://arxiv.org/pdf/1409.1556.pdf>`_.
 
     Args:
         features (nn.Layer): Vgg features create by function make_layers.
-        num_classes (int, optional): Output dim of last fc layer. If num_classes <=0, last fc layer 
+        num_classes (int, optional): Output dim of last fc layer. If num_classes <= 0, last fc layer 
                             will not be defined. Default: 1000.
         with_pool (bool, optional): Use pool before the last three fc layer or not. Default: True.
 
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of VGG model.
+
     Examples:
         .. code-block:: python
-          :name: code-example
 
             import paddle
             from paddle.vision.models import VGG
@@ -56,7 +58,6 @@ class VGG(nn.Layer):
 
             print(out.shape)
             # [1, 1000]
-
     """
 
     def __init__(self, features, num_classes=1000, with_pool=True):
@@ -139,15 +140,22 @@ def _vgg(arch, cfg, batch_norm, pretrained, **kwargs):
 
 
 def vgg11(pretrained=False, batch_norm=False, **kwargs):
-    """VGG 11-layer model
+    """VGG 11-layer model from
+    `"Very Deep Convolutional Networks For Large-Scale Image Recognition" <https://arxiv.org/pdf/1409.1556.pdf>`_.
     
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
-        batch_norm (bool): If True, returns a model with batch_norm layer. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        batch_norm (bool, optional): If True, returns a model with batch_norm layer. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`VGG <api_paddle_vision_VGG>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of VGG 11-layer model.
 
     Examples:
         .. code-block:: python
 
+            import paddle
             from paddle.vision.models import vgg11
 
             # build model
@@ -155,6 +163,12 @@ def vgg11(pretrained=False, batch_norm=False, **kwargs):
 
             # build vgg11 model with batch_norm
             model = vgg11(batch_norm=True)
+
+            x = paddle.rand([1, 3, 224, 224])
+            out = model(x)
+
+            print(out.shape)
+            # [1, 1000]
     """
     model_name = 'vgg11'
     if batch_norm:
@@ -163,15 +177,22 @@ def vgg11(pretrained=False, batch_norm=False, **kwargs):
 
 
 def vgg13(pretrained=False, batch_norm=False, **kwargs):
-    """VGG 13-layer model
+    """VGG 13-layer model from
+    `"Very Deep Convolutional Networks For Large-Scale Image Recognition" <https://arxiv.org/pdf/1409.1556.pdf>`_.
     
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
         batch_norm (bool): If True, returns a model with batch_norm layer. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`VGG <api_paddle_vision_VGG>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of VGG 13-layer model.
 
     Examples:
         .. code-block:: python
 
+            import paddle
             from paddle.vision.models import vgg13
 
             # build model
@@ -179,6 +200,12 @@ def vgg13(pretrained=False, batch_norm=False, **kwargs):
 
             # build vgg13 model with batch_norm
             model = vgg13(batch_norm=True)
+
+            x = paddle.rand([1, 3, 224, 224])
+            out = model(x)
+
+            print(out.shape)
+            # [1, 1000]
     """
     model_name = 'vgg13'
     if batch_norm:
@@ -187,15 +214,22 @@ def vgg13(pretrained=False, batch_norm=False, **kwargs):
 
 
 def vgg16(pretrained=False, batch_norm=False, **kwargs):
-    """VGG 16-layer model 
+    """VGG 16-layer model from
+    `"Very Deep Convolutional Networks For Large-Scale Image Recognition" <https://arxiv.org/pdf/1409.1556.pdf>`_.
     
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
-        batch_norm (bool): If True, returns a model with batch_norm layer. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        batch_norm (bool, optional): If True, returns a model with batch_norm layer. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`VGG <api_paddle_vision_VGG>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of VGG 16-layer model.
 
     Examples:
         .. code-block:: python
 
+            import paddle
             from paddle.vision.models import vgg16
 
             # build model
@@ -203,6 +237,12 @@ def vgg16(pretrained=False, batch_norm=False, **kwargs):
 
             # build vgg16 model with batch_norm
             model = vgg16(batch_norm=True)
+
+            x = paddle.rand([1, 3, 224, 224])
+            out = model(x)
+
+            print(out.shape)
+            # [1, 1000]
     """
     model_name = 'vgg16'
     if batch_norm:
@@ -211,15 +251,22 @@ def vgg16(pretrained=False, batch_norm=False, **kwargs):
 
 
 def vgg19(pretrained=False, batch_norm=False, **kwargs):
-    """VGG 19-layer model 
+    """VGG 19-layer model from
+    `"Very Deep Convolutional Networks For Large-Scale Image Recognition" <https://arxiv.org/pdf/1409.1556.pdf>`_.
     
     Args:
-        pretrained (bool): If True, returns a model pre-trained on ImageNet. Default: False.
-        batch_norm (bool): If True, returns a model with batch_norm layer. Default: False.
+        pretrained (bool, optional): Whether to load pre-trained weights. If True, returns a model pre-trained
+                            on ImageNet. Default: False.
+        batch_norm (bool, optional): If True, returns a model with batch_norm layer. Default: False.
+        **kwargs (optional): Additional keyword arguments. For details, please refer to :ref:`VGG <api_paddle_vision_VGG>`.
+
+    Returns:
+        :ref:`api_paddle_nn_Layer`. An instance of VGG 19-layer model.
 
     Examples:
         .. code-block:: python
 
+            import paddle
             from paddle.vision.models import vgg19
 
             # build model
@@ -227,6 +274,12 @@ def vgg19(pretrained=False, batch_norm=False, **kwargs):
 
             # build vgg19 model with batch_norm
             model = vgg19(batch_norm=True)
+
+            x = paddle.rand([1, 3, 224, 224])
+            out = model(x)
+
+            print(out.shape)
+            # [1, 1000]
     """
     model_name = 'vgg19'
     if batch_norm: