Modify L1DecayRegularizer, L1Decay, L2DecayRegularizer and L2Decay (#1217)

* modify L1DecayRegularizer, L1Decay, L2DecayRegularizer and L2Decay, test=develop

* add black lines, test=develop

doc/fluid/api_cn/regularizer_cn/L1DecayRegularizer_cn.rst

@@ -5,22 +5,22 @@ L1DecayRegularizer
 
 .. py:class:: paddle.fluid.regularizer.L1DecayRegularizer(regularization_coeff=0.0)
 
-实现 L1 权重衰减正则化。
-
-L1正则将会稀疏化权重矩阵。
+L1DecayRegularizer实现L1权重衰减正则化，用于模型训练，使得权重矩阵稀疏。
 
+具体实现中，L1权重衰减正则化的计算公式如下：
 
 .. math::
             \\L1WeightDecay=reg\_coeff∗sign(parameter)\\
 
-参数:
-  - **regularization_coeff** (float) – 正则化系数
+参数：
+  - **regularization_coeff** (float) – L1正则化系数，默认值为0.0。
 
 **代码示例**
 
 .. code-block:: python
 
     import paddle.fluid as fluid
+
     main_prog = fluid.Program()
     startup_prog = fluid.Program()
     with fluid.program_guard(main_prog, startup_prog):
@@ -35,11 +35,3 @@ L1正则将会稀疏化权重矩阵。
         regularization=fluid.regularizer.L1DecayRegularizer(
             regularization_coeff=0.1))
     optimizer.minimize(avg_loss)
-    
-  
-  
-
-
-
-
-

doc/fluid/api_cn/regularizer_cn/L1Decay_cn.rst

+
 .. _cn_api_fluid_regularizer_L1Decay:
 
 L1Decay
 -------------------------------
 
-.. py:attribute::   paddle.fluid.regularizer.L1Decay
+.. py:attribute::   paddle.fluid.regularizer.L1Decay(regularization_coeff=0.0)
+
+``L1Decay`` 是 ``L1DecayRegularizer`` 的别名。
+
+L1Decay实现L1权重衰减正则化，用于模型训练，使得权重矩阵稀疏。
+
+具体实现中，L1权重衰减正则化的计算公式如下：
+
+.. math::
+            \\L1WeightDecay=reg\_coeff∗sign(parameter)\\
+
+参数：
+  - **regularization_coeff** (float) – L1正则化系数，默认值为0.0。
+
+**代码示例**
+
+.. code-block:: python
+
+    import paddle.fluid as fluid
 
-``L1DecayRegularizer`` 的别名
+    main_prog = fluid.Program()
+    startup_prog = fluid.Program()
+    with fluid.program_guard(main_prog, startup_prog):
+        data = fluid.layers.data(name='image', shape=[3, 28, 28], dtype='float32')
+        label = fluid.layers.data(name='label', shape=[1], dtype='int64')
+        hidden = fluid.layers.fc(input=data, size=128, act='relu')
+        prediction = fluid.layers.fc(input=hidden, size=10, act='softmax')
+        loss = fluid.layers.cross_entropy(input=prediction, label=label)
+        avg_loss = fluid.layers.mean(loss)
+    optimizer = fluid.optimizer.Adagrad(
+        learning_rate=1e-4,
+        regularization=fluid.regularizer.L1Decay(
+            regularization_coeff=0.1))
+    optimizer.minimize(avg_loss)
 
 
 

doc/fluid/api_cn/regularizer_cn/L2DecayRegularizer_cn.rst

@@ -5,21 +5,22 @@ L2DecayRegularizer
 
 .. py:class:: paddle.fluid.regularizer.L2DecayRegularizer(regularization_coeff=0.0)
 
-实现L2 权重衰减正则化。 
+L2DecayRegularizer实现L2权重衰减正则化，用于模型训练，有助于防止模型对训练数据过拟合。
 
-较小的 L2 的有助于防止对训练数据的过度拟合。
+具体实现中，L2权重衰减正则化的计算公式如下：
 
 .. math::
             \\L2WeightDecay=reg\_coeff*parameter\\
 
 参数:
-  - **regularization_coeff** (float) – 正则化系数
+  - **regularization_coeff** (float) – 正则化系数，默认值为0.0。
   
 **代码示例**
 
 .. code-block:: python
 
     import paddle.fluid as fluid
+
     main_prog = fluid.Program()
     startup_prog = fluid.Program()
     with fluid.program_guard(main_prog, startup_prog):
@@ -34,10 +35,3 @@ L2DecayRegularizer
         regularization=fluid.regularizer.L2DecayRegularizer(
             regularization_coeff=0.1))
     optimizer.minimize(avg_loss)
-
-
-
-
-
-
-

doc/fluid/api_cn/regularizer_cn/L2Decay_cn.rst

@@ -5,9 +5,38 @@ L2Decay
 
 .. py:attribute::   paddle.fluid.regularizer.L2Decay
 
-``L2DecayRegularizer`` 的别名
+``L2Decay`` 是 ``L2DecayRegularizer`` 的别名。
 
+L2Decay实现L2权重衰减正则化，用于模型训练，有助于防止模型对训练数据过拟合。
 
+具体实现中，L2权重衰减正则化的计算公式如下：
+
+.. math::
+            \\L2WeightDecay=reg\_coeff*parameter\\
+
+参数:
+  - **regularization_coeff** (float) – 正则化系数，默认值为0.0。
+
+**代码示例**
+
+.. code-block:: python
+
+    import paddle.fluid as fluid
+
+    main_prog = fluid.Program()
+    startup_prog = fluid.Program()
+    with fluid.program_guard(main_prog, startup_prog):
+        data = fluid.layers.data(name='image', shape=[3, 28, 28], dtype='float32')
+        label = fluid.layers.data(name='label', shape=[1], dtype='int64')
+        hidden = fluid.layers.fc(input=data, size=128, act='relu')
+        prediction = fluid.layers.fc(input=hidden, size=10, act='softmax')
+        loss = fluid.layers.cross_entropy(input=prediction, label=label)
+        avg_loss = fluid.layers.mean(loss)
+    optimizer = fluid.optimizer.Adagrad(
+        learning_rate=1e-4,
+        regularization=fluid.regularizer.L2Decay(
+            regularization_coeff=0.1))
+    optimizer.minimize(avg_loss)