build_from_source.md 6.2 KB
Newer Older
L
liaogang 已提交
1
Installing from Sources
G
gangliao 已提交
2
==========================
Z
zhangjinchao01 已提交
3

L
liaogang 已提交
4 5 6
* [1. Download and Setup](#download)
* [2. Requirements](#requirements)
* [3. Build on Ubuntu](#ubuntu)
L
liaogang 已提交
7

L
liaogang 已提交
8
## <span id="download">Download and Setup</span> 
G
gangliao 已提交
9
You can download PaddlePaddle from the [github source](https://github.com/PaddlePaddle/Paddle).
Z
zhangjinchao01 已提交
10

L
liaogang 已提交
11
```bash
G
gangliao 已提交
12
git clone https://github.com/PaddlePaddle/Paddle paddle
13
cd paddle
L
Liu Yiqun 已提交
14
git submodule update --init --recursive
L
liaogang 已提交
15
```
Z
zhangjinchao01 已提交
16

L
liaogang 已提交
17
## <span id="requirements">Requirements</span>
Z
zhangjinchao01 已提交
18

19
To compile the source code, your computer must be equipped with GCC >=4.6 or Clang compiler.
L
liaogang 已提交
20
### Dependencies
Z
zhangjinchao01 已提交
21

L
liaogang 已提交
22 23 24 25
- **CMake**: version >= 2.8
- **BLAS**: MKL, OpenBlas or ATLAS
- **protobuf**: version >= 2.4, **Note: 3.x is not supported**
- **python**: only python 2.7 is supported currently
Z
zhangjinchao01 已提交
26

L
liaogang 已提交
27
### Options
Z
zhangjinchao01 已提交
28

L
liaogang 已提交
29
PaddlePaddle supports some build options. To enable it, first you need to install the related libraries. 
Z
zhangjinchao01 已提交
30

G
gangliao 已提交
31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
<html>
<table> 
<thead>
<tr>
<th scope="col" class="left">Optional</th>
<th scope="col" class="left">Description</th>
</tr>
</thead>
<tbody>
<tr><td class="left">WITH_GPU</td><td class="left">Compile with GPU mode.</td></tr>
<tr><td class="left">WITH_DOUBLE</td><td class="left">Compile with double precision floating-point, default: single precision.</td></tr>
<tr><td class="left">WITH_GLOG</td><td class="left">Compile with glog. If not found, default: an internal log implementation.</td></tr>
<tr><td class="left">WITH_GFLAGS</td><td class="left">Compile with gflags. If not found, default: an internal flag implementation.</td></tr>
<tr><td class="left">WITH_TESTING</td><td class="left">Compile with gtest for PaddlePaddle's unit testing.</td></tr>
<tr><td class="left">WITH_DOC</td><td class="left">	Compile to generate PaddlePaddle's docs, default: disabled (OFF).</td></tr>
<tr><td class="left">WITH_SWIG_PY</td><td class="left">Compile with python predict API, default: disabled (OFF).</td></tr>
<tr><td class="left">WITH_STYLE_CHECK</td><td class="left">Compile with code style check, default: enabled (ON).</td></tr>
</tbody>
G
gangliao 已提交
49
</table>
G
gangliao 已提交
50
</html>
Z
zhangjinchao01 已提交
51

L
liaogang 已提交
52 53 54 55
**Note:**
  - The GPU version works best with Cuda Toolkit 7.5 and cuDNN v5.
  - Other versions like Cuda Toolkit 6.5, 7.0, 8.0 and cuDNN v2, v3, v4 are also supported.
  - **To utilize cuDNN v5, Cuda Toolkit 7.5 is prerequisite and vice versa.**
Z
zhangjinchao01 已提交
56

L
liaogang 已提交
57
As a simple example, consider the following:  
Z
zhangjinchao01 已提交
58

L
liaogang 已提交
59
1. **Python Dependencies(optional)**
Z
zhangjinchao01 已提交
60
  
L
liaogang 已提交
61
    To compile PaddlePaddle with python predict API, make sure swig installed and set `-DWITH_SWIG_PY=ON` as follows:
Z
zhangjinchao01 已提交
62

L
liaogang 已提交
63 64 65 66 67
    ```bash
    # install swig on ubuntu
    sudo apt-get install swig
    # install swig on Mac OS X
    brew install swig
Z
zhangjinchao01 已提交
68

L
liaogang 已提交
69 70 71
    # active swig in cmake
    cmake .. -DWITH_SWIG_PY=ON
    ```
Z
zhangjinchao01 已提交
72

L
liaogang 已提交
73
2. **Doc Dependencies(optional)**
Z
zhangjinchao01 已提交
74

L
liaogang 已提交
75
    To generate PaddlePaddle's documentation, install dependencies and set `-DWITH_DOC=ON` as follows:
Z
zhangjinchao01 已提交
76

L
liaogang 已提交
77 78 79
    ```bash
    pip install 'sphinx>=1.4.0'
    pip install sphinx_rtd_theme breathe recommonmark
Z
zhangjinchao01 已提交
80

L
liaogang 已提交
81 82 83 84
    # install doxygen on Ubuntu
    sudo apt-get install doxygen 
    # install doxygen on Mac OS X
    brew install doxygen
Z
zhangjinchao01 已提交
85

L
liaogang 已提交
86 87 88
    # active docs in cmake
    cmake .. -DWITH_DOC=ON`
    ```
Z
zhangjinchao01 已提交
89

L
liaogang 已提交
90
## <span id="ubuntu">Build on Ubuntu 14.04</span>
Z
zhangjinchao01 已提交
91

L
liaogang 已提交
92
### Install Dependencies
Z
zhangjinchao01 已提交
93

L
liaogang 已提交
94
- **CPU Dependencies**
Z
zhangjinchao01 已提交
95

L
liaogang 已提交
96 97 98
    ```bash
    # necessary
    sudo apt-get update
L
liaogang 已提交
99
    sudo apt-get install -y g++ make cmake swig build-essential libatlas-base-dev python python-pip libpython-dev m4 libprotobuf-dev protobuf-compiler python-protobuf python-numpy git
L
liaogang 已提交
100 101 102 103 104 105 106 107 108 109 110 111 112
    # optional
    sudo apt-get install libgoogle-glog-dev
    sudo apt-get install libgflags-dev
    sudo apt-get install libgtest-dev
    sudo pip install wheel
    pushd /usr/src/gtest
    cmake .
    make
    sudo cp *.a /usr/lib
    popd
    ```
  
- **GPU Dependencies (optional)**
Z
zhangjinchao01 已提交
113

L
liaogang 已提交
114
    To build GPU version, you will need the following installed:
Z
zhangjinchao01 已提交
115

L
liaogang 已提交
116 117 118 119
        1. a CUDA-capable GPU
        2. A supported version of Linux with a gcc compiler and toolchain
        3. NVIDIA CUDA Toolkit (available at http://developer.nvidia.com/cuda-downloads)
        4. NVIDIA cuDNN Library (availabel at https://developer.nvidia.com/cudnn)
Z
zhangjinchao01 已提交
120

L
liaogang 已提交
121 122 123 124 125
    The CUDA development environment relies on tight integration with the host development environment,
    including the host compiler and C runtime libraries, and is therefore only supported on
    distribution versions that have been qualified for this CUDA Toolkit release.
        
    After downloading cuDNN library, issue the following commands:
Z
zhangjinchao01 已提交
126

L
liaogang 已提交
127 128 129 130
    ```bash
    sudo tar -xzf cudnn-7.5-linux-x64-v5.1.tgz -C /usr/local
    sudo chmod a+r /usr/local/cuda/include/cudnn.h /usr/local/cuda/lib64/libcudnn*
    ```
131
    Then you need to set LD\_LIBRARY\_PATH, PATH environment variables in ~/.bashrc.
L
liaogang 已提交
132 133 134 135 136 137 138 139 140

    ```bash
    export LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH
    export PATH=/usr/local/cuda/bin:$PATH
    ```

### Build and Install

As usual, the best option is to create build folder under paddle project directory.
Z
zhangjinchao01 已提交
141 142

```bash
L
liaogang 已提交
143 144
mkdir build && cd build
cmake ..
Z
zhangjinchao01 已提交
145 146
```

L
liaogang 已提交
147
CMake first check PaddlePaddle's dependencies in system default path. After installing some optional
L
liaogang 已提交
148 149
libraries, corresponding build option will be set automatically (for instance, glog, gtest and gflags).
If still not found, you can manually set it based on CMake error information from your screen.
Z
zhangjinchao01 已提交
150

L
liaogang 已提交
151 152
As a simple example, consider the following:

L
liaogang 已提交
153
- **Only CPU with swig**
L
liaogang 已提交
154 155

  ```bash
L
liaogang 已提交
156
  cmake  .. -DWITH_GPU=OFF -DWITH_SWIG_PY=ON
L
liaogang 已提交
157
  ```
L
liaogang 已提交
158
- **GPU with swig**
L
liaogang 已提交
159 160

  ```bash
L
liaogang 已提交
161
  cmake .. -DWITH_GPU=ON -DWITH_SWIG_PY=ON
L
liaogang 已提交
162 163 164 165 166 167 168
  ```

- **GPU with doc and swig**

  ```bash
  cmake .. -DWITH_GPU=ON -DWITH_DOC=ON -DWITH_SWIG_PY=ON
  ``` 
Z
zhangjinchao01 已提交
169

170
Finally, you can build PaddlePaddle:
Z
zhangjinchao01 已提交
171 172 173

```bash
# you can add build option here, such as:    
L
liaogang 已提交
174
cmake .. -DWITH_GPU=ON -DCMAKE_INSTALL_PREFIX=<path to install> -DWITH_SWIG_PY=ON
175
# please use sudo make install, if you want to install PaddlePaddle into the system
Z
zhangjinchao01 已提交
176
make -j `nproc` && make install
L
liaogang 已提交
177
# set PaddlePaddle installation path in ~/.bashrc
L
liaogang 已提交
178
export PATH=<path to install>/bin:$PATH
Z
zhangjinchao01 已提交
179 180
```

L
liaogang 已提交
181 182 183 184
If you set `WITH_SWIG_PY=ON`, related python dependencies also need to be installed.
Otherwise, PaddlePaddle will automatically install python dependencies
at first time when user run paddle commands, such as `paddle version`, `paddle train`.
It may require sudo privileges:
Z
zhangjinchao01 已提交
185 186

```bash
L
liaogang 已提交
187 188 189 190
# you can run
sudo pip install <path to install>/opt/paddle/share/wheels/*.whl
# or just run 
sudo paddle version
Z
zhangjinchao01 已提交
191
```