diff --git a/doc/tutorials/image_classification/index_cn.md b/doc/tutorials/image_classification/index_cn.md
new file mode 100644
index 0000000000000000000000000000000000000000..87f465522a0fa21c8c03754b4be8dcb035c4de81
--- /dev/null
+++ b/doc/tutorials/image_classification/index_cn.md
@@ -0,0 +1,205 @@
+图像分类教程
+==========
+
+在本教程中，我们将使用CIFAR-10数据集训练一个卷积神经网络，并使用这个神经网络来对图片进行分类。如下图所示，卷积神经网络可以辨识图片中的主体，并给出分类结果。
+<center>![Image Classification](./image_classification.png)</center>
+
+## 数据准备
+首先下载CIFAR-10数据集。下面是CIFAR-10数据集的官方网址：
+
+<https://www.cs.toronto.edu/~kriz/cifar.html>
+
+我们准备了一个脚本，可以用于从官方网站上下载CIFAR-10数据集，转为jpeg文件并存入特定的目录。使用这个脚本前请确认已经安装了pillow及相关依赖模块。可以参照下面的命令进行安装：
+
+1. 安装pillow
+
+```bash
+sudo apt-get install libjpeg-dev
+pip install pillow
+```
+
+2. 下载数据集
+
+```bash
+cd demo/image_classification/data/
+sh download_cifar.sh
+```
+
+CIFAR-10数据集包含60000张32x32的彩色图片。图片分为10类，每个类包含6000张。其中50000张图片作为训练集，10000张作为测试集。
+
+下图展示了所有的图片类别，每个类别中随机抽取了10张图片。
+<center>![Image Classification](./cifar.png)</center>
+
+脚本运行完成后，我们应当会得到一个名为cifar-out的文件夹，其下子文件夹的结构如下
+
+
+```
+train
+---airplane
+---automobile
+---bird
+---cat
+---deer
+---dog
+---frog
+---horse
+---ship
+---truck
+test
+---airplane
+---automobile
+---bird
+---cat
+---deer
+---dog
+---frog
+---horse
+---ship
+---truck
+```
+
+cifar-out下包含`train`和`test`两个文件夹，其中分别包含了CIFAR-10中的训练集和测试集。这两个文件夹下各自有10个子文件夹，每个子文件夹下存储相应分类的图片。将图片按照上述结构存储好之后，我们就可以着手对分类模型进行训练了。
+
+## 预处理
+数据下载之后，还需要进行预处理，将数据转换为Paddle的格式。我们可以通过如下命令进行预处理工作：
+
+```
+cd demo/image_classification/
+sh preprocess.sh
+```
+
+其中`preprocess.sh` 调用 `./demo/image_classification/preprocess.py` 对图片进行预处理
+```sh
+export PYTHONPATH=$PYTHONPATH:../../
+data_dir=./data/cifar-out
+python preprocess.py -i $data_dir -s 32 -c 1
+```
+
+`./demo/image_classification/preprocess.py` 使用如下参数：
+
+- `-i` 或 `--input` 给出输入数据所在路径；
+- `-s` 或 `--size` 给出图片尺寸；
+- `-c` 或 `--color` 标示图片是彩色图或灰度图
+
+## 模型训练
+在开始训练之前，我们需要先创建一个模型配置文件。下面我们给出了一个配置示例。**注意**，这里的列出的和`vgg_16_cifar.py`文件稍有差别，因为该文件可适用于预测。
+
+```python
+from paddle.trainer_config_helpers import *
+data_dir='data/cifar-out/batches/'
+meta_path=data_dir+'batches.meta'
+args = {'meta':meta_path, 'mean_img_size': 32,
+        'img_size': 32, 'num_classes': 10,
+        'use_jpeg': 1, 'color': "color"}
+define_py_data_sources2(train_list=data_dir+"train.list",
+                        test_list=data_dir+'test.list',
+                        module='image_provider',
+                        obj='processData',
+                        args=args)
+settings(
+    batch_size = 128,
+    learning_rate = 0.1 / 128.0,
+    learning_method = MomentumOptimizer(0.9),
+    regularization = L2Regularization(0.0005 * 128))
+
+img = data_layer(name='image', size=3*32*32)
+lbl = data_layer(name="label", size=10)
+# small_vgg is predined in trainer_config_helpers.network
+predict = small_vgg(input_image=img, num_channels=3)
+outputs(classification_cost(input=predict, label=lbl))
+```
+
+在第一行中我们载入用于定义网络的函数。
+```python
+from paddle.trainer_config_helpers import *
+```
+
+之后定义的`define_py_data_sources2`使用Python数据提供器，其中 `args`将在`image_provider.py`进行使用，该文件负责产生图片数据并传递给Paddle系统
+ - `meta`: 训练集平均值。
+ - `mean_img_size`: 平均特征图的高度及宽度。
+ - `img_size`：输入图片的高度及宽度。
+ - `num_classes`：类别个数。
+ - `use_jpeg`：处理过程中数据存储格式。
+ - `color`：标示是否为彩色图片。
+ 
+ `settings`用于设置训练算法。在下面的例子中，learning rate被设置为0.1除以batch size，而weight decay则为0.0005乘以batch size。
+ 
+ ```python
+settings(
+    batch_size = 128,
+    learning_rate = 0.1 / 128.0,
+    learning_method = MomentumOptimizer(0.9),
+    regularization = L2Regularization(0.0005 * 128)
+)
+```
+
+`small_vgg`定义了网络结构。这里我们使用的是一个小的VGG网络。关于VGG卷积神经网络的描述可以参考：[http://www.robots.ox.ac.uk/~vgg/research/very_deep/](http://www.robots.ox.ac.uk/~vgg/research/very_deep/)。
+```python
+# small_vgg is predined in trainer_config_helpers.network
+predict = small_vgg(input_image=img, num_channels=3)
+```
+配置创建完毕后，可以运行脚本train.sh来训练模型。
+
+```bash
+config=vgg_16_cifar.py
+output=./cifar_vgg_model
+log=train.log
+
+paddle train \
+--config=$config \
+--dot_period=10 \
+--log_period=100 \
+--test_all_data_in_one_period=1 \
+--use_gpu=1 \
+--save_dir=$output \
+2>&1 | tee $log
+
+python -m paddle.utils.plotcurve -i $log > plot.png
+```
+- 这里我们使用的是GPU模式进行训练。如果你没有GPU环境，可以设置`use_gpu=0`。
+- `./demo/image_classification/vgg_16_cifar.py`是网络和数据配置文件。各项参数的详细说明可以在命令行参数相关文档中找到。
+- 脚本`plotcurve.py`依赖于python的`matplotlib`模块。因此如果这个脚本运行失败，也许是因为需要安装`matplotlib`。
+在训练完成后，训练及测试误差曲线图会被`plotcurve.py`脚本保存在 `plot.png`中。下面是一个误差曲线图的示例：
+
+<center>![Training and testing curves.](./plot.png)</center>
+
+## 预测
+在训练完成后，模型及参数会被保存在路径`./cifar_vgg_model/pass-%05d`下。例如第300个pass的模型会被保存在`./cifar_vgg_model/pass-00299`。
+
+要对一个图片的进行分类预测，我们可以使用`predict.sh`，该脚本将输出预测分类的标签：
+
+```
+sh predict.sh
+```
+
+predict.sh:
+```
+model=cifar_vgg_model/pass-00299/
+image=data/cifar-out/test/airplane/seaplane_s_000978.png
+use_gpu=1
+python prediction.py $model $image $use_gpu
+```
+
+## 练习
+在CUB-200数据集上使用VGG模型训练一个鸟类图片分类模型。相关的鸟类数据集可以从如下地址下载，其中包含了200种鸟类的照片（主要来自北美洲）。
+
+<http://www.vision.caltech.edu/visipedia/CUB-200.html>
+
+
+
+
+## 细节探究
+### 卷积神经网络
+卷积神经网络是一种使用卷积层的前向神经网络，很适合构建用于理解图片内容的模型。一个典型的神经网络如下图所示：
+
+![Convolutional Neural Network](./lenet.png)
+
+一个卷积神经网络包含如下层：
+
+- 卷积层：通过卷积操作从图片或特征图中提取特征
+- 池化层：使用max-pooling对特征图下采样
+- 全连接层：使输入层到隐藏层的神经元是全部连接的。
+
+卷积神经网络在图片分类上有着惊人的性能，这是因为它发掘出了图片的两类重要信息：局部关联性质和空间不变性质。通过交替使用卷积和池化处理， 卷积神经网络能够很好的表示这两类信息。
+
+关于如何定义网络中的层，以及如何在层之间进行连接，请参考Layer文档。
diff --git a/doc/tutorials/image_classification/index_en.md b/doc/tutorials/image_classification/index_en.md
index 29cfc99702c362d1eaeeff5332f56122b8de337a..60c81a6a539944634773f38ec4c9a59709dd4afc 100644
--- a/doc/tutorials/image_classification/index_en.md
+++ b/doc/tutorials/image_classification/index_en.md
@@ -147,7 +147,7 @@ for classification. A description of VGG network can be found here [http://www.r
 # small_vgg is predined in trainer_config_helpers.network
 predict = small_vgg(input_image=img, num_channels=3)
 ```
-After writing the config, we can train the model by running the script train.sh. Notice that the following script assumes the you run the script in the `./demo/image_classification` folder. If you run the script in a different folder, you need to change the paths of the scripts and the configuration files accordingly.
+After writing the config, we can train the model by running the script train.sh.
 
 ```bash
 config=vgg_16_cifar.py
diff --git a/paddle/api/PaddleAPI.h b/paddle/api/PaddleAPI.h
index 84a66719c33678fc4aeb038bb81a6b7c5d0c93fb..5c4c25e770facc734314dde4118089061a2cc6ff 100644
--- a/paddle/api/PaddleAPI.h
+++ b/paddle/api/PaddleAPI.h
@@ -20,15 +20,11 @@ limitations under the License. */
 #include <string>
 #include <vector>
 #include "paddle/utils/GlobalConstants.h"
-#include "paddle/utils/TypeDefs.h"
+#include "paddle/utils/common.h"
 
 /// Import PaddlePaddle's enumeration into global namespace.
 using namespace paddle::enumeration_wrapper;  // NOLINT
 
-#define DISABLE_COPY_AND_ASSIGN(classname) \
-  classname(const classname& other);       \
-  classname& operator=(const classname& other)
-
 /**
  * @brief Initialize paddle.
  *
@@ -102,7 +98,7 @@ const size_t NO_SPARSE_ID = -1UL;
 struct MatrixPrivate;
 class Matrix {
   Matrix();  // User Cannot Create Matrix.
-  DISABLE_COPY_AND_ASSIGN(Matrix);
+  DISABLE_COPY(Matrix);
   static Matrix* createByPaddleMatrixPtr(void* sharedPtr);
 
 public:
@@ -242,7 +238,7 @@ private:
 
 struct VectorPrivate;
 class Vector {
-  DISABLE_COPY_AND_ASSIGN(Vector);
+  DISABLE_COPY(Vector);
   Vector();
   static Vector* createByPaddleVectorPtr(void* ptr);
 
@@ -322,7 +318,7 @@ private:
 struct IVectorPrivate;
 class IVector {
   IVector();
-  DISABLE_COPY_AND_ASSIGN(IVector);
+  DISABLE_COPY(IVector);
   static IVector* createByPaddleVectorPtr(void* ptr);
 
 public:
@@ -402,7 +398,7 @@ struct ArgumentsPrivate;
 class Arguments {
 private:
   Arguments();  // Internal Create.
-  DISABLE_COPY_AND_ASSIGN(Arguments);
+  DISABLE_COPY(Arguments);
 
 public:
   /**
@@ -472,7 +468,7 @@ enum GradientMatchineCreateMode {
 
 struct ParameterConfigPrivate;
 class ParameterConfig {
-  DISABLE_COPY_AND_ASSIGN(ParameterConfig);
+  DISABLE_COPY(ParameterConfig);
   ParameterConfig();
 
   /**
@@ -502,7 +498,7 @@ private:
 
 struct OptimizationConfigPrivate;
 class OptimizationConfig {
-  DISABLE_COPY_AND_ASSIGN(OptimizationConfig);
+  DISABLE_COPY(OptimizationConfig);
   OptimizationConfig();
 
 public:
@@ -526,7 +522,7 @@ struct ParameterPrivate;
 class Parameter {
 private:
   Parameter();
-  DISABLE_COPY_AND_ASSIGN(Parameter);
+  DISABLE_COPY(Parameter);
 
 public:
   virtual ~Parameter();
@@ -568,7 +564,7 @@ struct ModelConfigPrivate;
 class ModelConfig {
 private:
   ModelConfig();
-  DISABLE_COPY_AND_ASSIGN(ModelConfig);
+  DISABLE_COPY(ModelConfig);
 
 public:
   virtual ~ModelConfig();
@@ -589,7 +585,7 @@ struct TrainerConfigPrivate;
 class TrainerConfig {
 private:
   TrainerConfig();
-  DISABLE_COPY_AND_ASSIGN(TrainerConfig);
+  DISABLE_COPY(TrainerConfig);
 
 public:
   virtual ~TrainerConfig();
@@ -629,7 +625,7 @@ public:
 
 struct ParameterTraverseCallbackPrivate;
 class ParameterTraverseCallback {
-  DISABLE_COPY_AND_ASSIGN(ParameterTraverseCallback);
+  DISABLE_COPY(ParameterTraverseCallback);
   ParameterTraverseCallback();
 
 public:
@@ -651,7 +647,7 @@ private:
  */
 struct ParameterOptimizerPrivate;
 class ParameterOptimizer {
-  DISABLE_COPY_AND_ASSIGN(ParameterOptimizer);
+  DISABLE_COPY(ParameterOptimizer);
   ParameterOptimizer();
 
 public:
@@ -688,7 +684,7 @@ struct GradientMachinePrivate;
 class GradientMachine {
 private:
   GradientMachine();
-  DISABLE_COPY_AND_ASSIGN(GradientMachine);
+  DISABLE_COPY(GradientMachine);
 
 public:
   virtual ~GradientMachine();
@@ -780,7 +776,7 @@ private:
   TrainerPrivate* m;
   Trainer();
   Trainer(TrainerConfig* optConfig, GradientMachine* gm);
-  DISABLE_COPY_AND_ASSIGN(Trainer);
+  DISABLE_COPY(Trainer);
 
 public:
   virtual ~Trainer();
@@ -846,7 +842,7 @@ public:
 
 struct SequenceGeneratorPrivate;
 class SequenceGenerator {
-  DISABLE_COPY_AND_ASSIGN(SequenceGenerator);
+  DISABLE_COPY(SequenceGenerator);
   SequenceGenerator();
 
 public:
diff --git a/paddle/cuda/include/hl_base.h b/paddle/cuda/include/hl_base.h
index 84c5f2d5c91feb7896643d2c5f60a279ebe944e7..5b9884b786530aee91312c7547496c94cd7a89cb 100644
--- a/paddle/cuda/include/hl_base.h
+++ b/paddle/cuda/include/hl_base.h
@@ -16,7 +16,31 @@ limitations under the License. */
 #define HL_BASE_H_
 
 #include <cstddef>
-#include "paddle/utils/TypeDefs.h"
+
+#ifdef PADDLE_TYPE_DOUBLE
+#define HL_FLOAT_MAX 3.40282347e+38F
+#define HL_FLOAT_MIN 1.17549435e-38F
+using real = double;
+#else
+#define HL_FLOAT_MAX 1.7976931348623157e+308
+#define HL_FLOAT_MIN 2.2250738585072014e-308
+using real = float;
+#endif
+
+/**
+ * The maximum input value for exp, used to avoid overflow problem.
+ * currently only used for tanh function.
+ */
+#define EXP_MAX_INPUT 40.0
+
+/**
+ * @brief DIVUP(x, y) is similar to ceil(x / y).
+ * @note  For CUDA, DIVUP will be used to specify
+ *        the size of blockDim.
+ */
+#ifndef DIVUP
+#define DIVUP(x, y) (((x) + (y)-1) / (y))
+#endif
 
 /**
  * HPPL is an internal high performance parallel computing library
@@ -181,46 +205,6 @@ typedef struct {
   size_t nnz;
 } _hl_sparse_matrix_s, *hl_sparse_matrix_s;
 
-#ifndef PADDLE_TYPE_DOUBLE
-/**
- * HPPL data type: real (float or double)
- *
- * if real == float
- *
- * HL_FLOAT_MAX: 3.40282347e+38F
- *
- * HL_FLOAT_MIN: 1.17549435e-38F
- */
-#define HL_FLOAT_MAX 3.40282347e+38F
-/**
- * if real == double
- *
- * HL_FLOAT_MAX: 1.7976931348623157e+308
- *
- * HL_FLOAT_MIN: 2.2250738585072014e-308
- */
-#define HL_FLOAT_MIN 1.17549435e-38F
-#else
-#define HL_FLOAT_MAX 1.7976931348623157e+308
-#define HL_FLOAT_MIN 2.2250738585072014e-308
-#endif
-
-/**
- * The maximum input value for exp, used to avoid overflow problem.
- *
- * Currently only used for tanh function.
- */
-#define EXP_MAX_INPUT 40.0
-
-/**
- * @brief DIVUP(x, y) is similar to ceil(x / y).
- * @note  For CUDA, DIVUP will be used to specify
- *        the size of blockDim.
- */
-#ifndef DIVUP
-#define DIVUP(x, y) (((x) + (y)-1) / (y))
-#endif
-
 #ifdef __NVCC__
 
 #include "cuda_runtime.h"
diff --git a/paddle/gserver/dataproviders/DataProvider.h b/paddle/gserver/dataproviders/DataProvider.h
index 9b7f7e36cedaa230ae0694d87cc033bd6fa6e652..5f031fc7c0761a8fe97eb16fe1dd8e0a1debfcdb 100644
--- a/paddle/gserver/dataproviders/DataProvider.h
+++ b/paddle/gserver/dataproviders/DataProvider.h
@@ -34,8 +34,8 @@ limitations under the License. */
 #include "paddle/utils/Logging.h"
 #include "paddle/utils/Queue.h"
 #include "paddle/utils/ThreadLocal.h"
-#include "paddle/utils/TypeDefs.h"
 #include "paddle/utils/Util.h"
+#include "paddle/utils/common.h"
 
 namespace paddle {
 /**
diff --git a/paddle/gserver/layers/BatchNormalizationLayer.h b/paddle/gserver/layers/BatchNormalizationLayer.h
index 052c2077322be59f9d41966c1c8b6ab20c8f85bb..195acbbfc58db8368f6db1c1595dd6b04801ee26 100644
--- a/paddle/gserver/layers/BatchNormalizationLayer.h
+++ b/paddle/gserver/layers/BatchNormalizationLayer.h
@@ -58,6 +58,8 @@ protected:
   /// to batch, channels* imagePixels.
   void shrinkMat(const MatrixPtr& in, MatrixPtr& out);
 
+  void onPassEnd() { firstTest_ = true; }
+
   MatrixPtr tmpMat_, tmpGrad_;
   MatrixPtr expandedIn_, expandedOut_;
   MatrixPtr expandedInGrad_, expandedOutGrad_, inGrad_;
diff --git a/paddle/gserver/layers/GruCompute.h b/paddle/gserver/layers/GruCompute.h
index 42c0019319ac9f20f9c3349fb2429c30f03d682b..a56af21317d1d43c836f7fe599a4dc614804bfec 100644
--- a/paddle/gserver/layers/GruCompute.h
+++ b/paddle/gserver/layers/GruCompute.h
@@ -16,7 +16,7 @@ limitations under the License. */
 
 #include "ModelConfig.pb.h"
 #include "hl_gpu.h"
-#include "paddle/utils/TypeDefs.h"
+#include "paddle/utils/common.h"
 
 namespace paddle {
 
diff --git a/paddle/gserver/layers/LstmCompute.h b/paddle/gserver/layers/LstmCompute.h
index 140a4c6ecf5cfaf1045cec3ca2db5d4f2e54aca4..0d65b4158ebdc04f199048bbba98317c89fc8beb 100644
--- a/paddle/gserver/layers/LstmCompute.h
+++ b/paddle/gserver/layers/LstmCompute.h
@@ -16,7 +16,7 @@ limitations under the License. */
 
 #include "ModelConfig.pb.h"
 #include "hl_gpu.h"
-#include "paddle/utils/TypeDefs.h"
+#include "paddle/utils/common.h"
 
 namespace paddle {
 
diff --git a/paddle/gserver/layers/MultinomialSampler.h b/paddle/gserver/layers/MultinomialSampler.h
index 677b047029305549084770bdb5eadfeaafbfac8a..b48073c80b6f57cd86ceb80b9d749548c3acc1ac 100644
--- a/paddle/gserver/layers/MultinomialSampler.h
+++ b/paddle/gserver/layers/MultinomialSampler.h
@@ -16,7 +16,7 @@ limitations under the License. */
 
 #include <memory>
 #include <random>
-#include "paddle/utils/TypeDefs.h"
+#include "paddle/utils/common.h"
 
 namespace paddle {
 
diff --git a/paddle/math/BaseMatrix.h b/paddle/math/BaseMatrix.h
index 2933c20fbad930248c41969d88d45cf397b9dcf8..8f9bc9e823eb8062535920361899ce3cc06ec3a7 100644
--- a/paddle/math/BaseMatrix.h
+++ b/paddle/math/BaseMatrix.h
@@ -16,7 +16,7 @@ limitations under the License. */
 #include <stdint.h>
 #include <cstddef>
 #include "TensorExpression.h"
-#include "paddle/utils/TypeDefs.h"
+#include "paddle/utils/common.h"
 
 namespace paddle {
 
diff --git a/paddle/math/Matrix.h b/paddle/math/Matrix.h
index 25ce09e346694298e4901e52ab1ec6a3a8044263..bda863de38675fe481544a7e82b69f445df361bd 100644
--- a/paddle/math/Matrix.h
+++ b/paddle/math/Matrix.h
@@ -27,7 +27,7 @@ limitations under the License. */
 #include "MemoryHandle.h"
 #include "Vector.h"
 #include "paddle/utils/ThreadLocal.h"
-#include "paddle/utils/TypeDefs.h"
+#include "paddle/utils/common.h"
 
 namespace paddle {
 
diff --git a/paddle/math/TensorExpression.h b/paddle/math/TensorExpression.h
index 9bd789e8c511f33d8415e421281e99eb10fc63fe..f3d60e400380f7d7d645559318837b0d7706661d 100644
--- a/paddle/math/TensorExpression.h
+++ b/paddle/math/TensorExpression.h
@@ -17,7 +17,7 @@ limitations under the License. */
 #include <cstddef>
 #include "hl_tensor_ops.h"
 #include "paddle/utils/Logging.h"
-#include "paddle/utils/TypeDefs.h"
+#include "paddle/utils/common.h"
 
 namespace paddle {
 
diff --git a/paddle/math/Vector.h b/paddle/math/Vector.h
index 8a24103bd4107035c8068c24ec3be6ec06957112..b4347a70f874a2a1bf933bbea4d1b15385f36090 100644
--- a/paddle/math/Vector.h
+++ b/paddle/math/Vector.h
@@ -22,7 +22,7 @@ limitations under the License. */
 #include "BaseMatrix.h"
 #include "MemoryHandle.h"
 #include "paddle/utils/Thread.h"
-#include "paddle/utils/TypeDefs.h"
+#include "paddle/utils/common.h"
 
 namespace paddle {
 
diff --git a/paddle/parameter/ParallelParameter.h b/paddle/parameter/ParallelParameter.h
index 417e386dc74d308a6c0aefa2640f0f37de8dbf1f..1ee220d2dc1a26b3f394ca673975cc827f450206 100644
--- a/paddle/parameter/ParallelParameter.h
+++ b/paddle/parameter/ParallelParameter.h
@@ -28,7 +28,7 @@ limitations under the License. */
 #include "paddle/parameter/ParameterUpdateFunctions.h"
 #include "paddle/utils/Flags.h"
 #include "paddle/utils/Locks.h"
-#include "paddle/utils/TypeDefs.h"
+#include "paddle/utils/common.h"
 
 #include "ParameterConfig.pb.h"
 
diff --git a/paddle/parameter/Parameter.h b/paddle/parameter/Parameter.h
index 532c6770e596c33dfe7fd42f32157b2c6c19e18e..e05137b315f254795de26a5ff0ac977e7968f4d8 100644
--- a/paddle/parameter/Parameter.h
+++ b/paddle/parameter/Parameter.h
@@ -29,8 +29,8 @@ limitations under the License. */
 #include "paddle/utils/GlobalConstants.h"
 #include "paddle/utils/Locks.h"
 #include "paddle/utils/ThreadLocal.h"
-#include "paddle/utils/TypeDefs.h"
 #include "paddle/utils/Util.h"
+#include "paddle/utils/common.h"
 
 namespace paddle {
 
diff --git a/paddle/parameter/ParameterUpdateFunctions.h b/paddle/parameter/ParameterUpdateFunctions.h
index 2d277e47e7eafc118fa37343e93e8a331a260aa9..2cb379871716ffd9e75eede607276b6b3f200e6b 100644
--- a/paddle/parameter/ParameterUpdateFunctions.h
+++ b/paddle/parameter/ParameterUpdateFunctions.h
@@ -15,7 +15,7 @@ limitations under the License. */
 #pragma once
 
 #include "paddle/math/Vector.h"
-#include "paddle/utils/TypeDefs.h"
+#include "paddle/utils/common.h"
 
 namespace paddle {
 
diff --git a/paddle/pserver/BaseClient.h b/paddle/pserver/BaseClient.h
index 262afafbe2d61305a158d945fac2d3b265012cbd..ccf05ae1ca3ab76fbe9d36237969207768de4dd2 100644
--- a/paddle/pserver/BaseClient.h
+++ b/paddle/pserver/BaseClient.h
@@ -18,7 +18,7 @@ limitations under the License. */
 #include "paddle/math/Matrix.h"
 #include "paddle/pserver/ProtoServer.h"
 #include "paddle/utils/Queue.h"
-#include "paddle/utils/TypeDefs.h"
+#include "paddle/utils/common.h"
 
 namespace paddle {
 
diff --git a/paddle/pserver/ParameterClient2.h b/paddle/pserver/ParameterClient2.h
index eed71ccb43b0fec76a74a7f00662c32c97c26ff4..70cfc6d70072f399ef97eef1a0e6111a127cbd9f 100644
--- a/paddle/pserver/ParameterClient2.h
+++ b/paddle/pserver/ParameterClient2.h
@@ -26,8 +26,8 @@ limitations under the License. */
 #include "paddle/utils/Flags.h"
 #include "paddle/utils/Locks.h"
 #include "paddle/utils/Queue.h"
-#include "paddle/utils/TypeDefs.h"
 #include "paddle/utils/Util.h"
+#include "paddle/utils/common.h"
 
 #include "ParameterService.pb.h"
 
diff --git a/paddle/pserver/ParameterServer2.h b/paddle/pserver/ParameterServer2.h
index b0cf22e1fb158e76fcee1ce6ef1f375995803ce6..79d1eb97ff149f4f5ca9a924c1b0b7ba629f1e33 100644
--- a/paddle/pserver/ParameterServer2.h
+++ b/paddle/pserver/ParameterServer2.h
@@ -32,7 +32,7 @@ limitations under the License. */
 #include "paddle/utils/Locks.h"
 #include "paddle/utils/Stat.h"
 #include "paddle/utils/ThreadLocal.h"
-#include "paddle/utils/TypeDefs.h"
+#include "paddle/utils/common.h"
 
 #include "ParameterService.pb.h"
 
diff --git a/paddle/utils/CpuId.h b/paddle/utils/CpuId.h
index 7a354da75851ed7cca4e85e77714624634951f00..1218e8194c4e837ca880744f92e769a68ba474de 100644
--- a/paddle/utils/CpuId.h
+++ b/paddle/utils/CpuId.h
@@ -11,7 +11,7 @@ limitations under the License. */
 
 #pragma once
 
-#include "DisableCopy.h"
+#include "common.h"
 
 namespace paddle {
 
diff --git a/paddle/utils/DisableCopy.h b/paddle/utils/DisableCopy.h
deleted file mode 100644
index 41de98bbde664651803c8db4c0cd7216b2ff4231..0000000000000000000000000000000000000000
--- a/paddle/utils/DisableCopy.h
+++ /dev/null
@@ -1,23 +0,0 @@
-/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserve.
-
-Licensed under the Apache License, Version 2.0 (the "License");
-you may not use this file except in compliance with the License.
-You may obtain a copy of the License at
-
-    http://www.apache.org/licenses/LICENSE-2.0
-
-Unless required by applicable law or agreed to in writing, software
-distributed under the License is distributed on an "AS IS" BASIS,
-WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-See the License for the specific language governing permissions and
-limitations under the License. */
-
-#pragma once
-
-/**
- * Disable copy macro.
- */
-#define DISABLE_COPY(CLASS_NAME)                \
-  CLASS_NAME(CLASS_NAME &&) = delete;           \
-  CLASS_NAME(const CLASS_NAME &other) = delete; \
-  CLASS_NAME &operator=(const CLASS_NAME &other) = delete
diff --git a/paddle/utils/Locks.h b/paddle/utils/Locks.h
index 0f922f3548d97eb16ca897564faf1bf083f0d5ac..a21872e89ebc172b87c8b5c3731a89302f34f521 100644
--- a/paddle/utils/Locks.h
+++ b/paddle/utils/Locks.h
@@ -19,7 +19,7 @@ limitations under the License. */
 #include <condition_variable>
 #include <mutex>
 
-#include "DisableCopy.h"
+#include "common.h"
 
 namespace paddle {
 
diff --git a/paddle/utils/Util.h b/paddle/utils/Util.h
index e5a89070f1a953d70a43321cb5417656c907ee9d..dc15ada5862d648af27aa1b0e8c8a5cce012ded8 100644
--- a/paddle/utils/Util.h
+++ b/paddle/utils/Util.h
@@ -26,12 +26,11 @@ limitations under the License. */
 #include <unordered_map>
 #include <vector>
 
-#include "DisableCopy.h"
 #include "Logging.h"
 #include "TrainerConfig.pb.h"
+#include "common.h"
 
 #include "Flags.h"
-#include "TypeDefs.h"
 #include "hl_gpu.h"
 
 /**
diff --git a/paddle/utils/Version.h b/paddle/utils/Version.h
index d1a07d9485076e5382d47f7408fcbf032166b1ed..aa5df3243893145dbcc7e7ef2592555fc1c88fc9 100644
--- a/paddle/utils/Version.h
+++ b/paddle/utils/Version.h
@@ -15,7 +15,7 @@ limitations under the License. */
 #pragma once
 #include <stddef.h>
 #include <iostream>
-#include "TypeDefs.h"
+#include "common.h"
 
 namespace paddle {
 
diff --git a/paddle/utils/TypeDefs.h b/paddle/utils/common.h
similarity index 71%
rename from paddle/utils/TypeDefs.h
rename to paddle/utils/common.h
index c50a05e82daefd1273c896f3603957f4484ecd5d..3ff0b869478832d023956d190ddeb89a69acf58b 100644
--- a/paddle/utils/TypeDefs.h
+++ b/paddle/utils/common.h
@@ -15,12 +15,19 @@ limitations under the License. */
 #pragma once
 
 namespace paddle {
+
+/**
+ * Disable copy macro.
+ */
+#define DISABLE_COPY(class_name)                \
+  class_name(class_name &&) = delete;           \
+  class_name(const class_name &other) = delete; \
+  class_name &operator=(const class_name &other) = delete
+
 #ifdef PADDLE_TYPE_DOUBLE
-typedef double real;
+using real = double;
 #else
-typedef float real;
+using real = float;
 #endif
 
 }  // namespace paddle
-
-using paddle::real;