Skip to content

P
Paddle

BaiXuePrincess / Paddle
与 Fork 源项目一致

Fork自 PaddlePaddle / Paddle

通知 1
Star 1
Fork 0
P
Paddle
切换分支/标签
查找文件 普通视图历史永久链接Permalink
cpu_profiling.html 30.7 KB
Newer Older
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
1 2 3 4 5 6 7 8 9 10 


<!DOCTYPE html>
<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
<head>
  <meta charset="utf-8">
  
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
11 
  <title>Profiling the Python Code &mdash; PaddlePaddle  documentation</title>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 
  

  
  

  

  
  
    

  

  
  
    <link rel="stylesheet" href="../../_static/css/theme.css" type="text/css" />
  

  
  
        <link rel="index" title="Index"
              href="../../genindex.html"/>
        <link rel="search" title="Search" href="../../search.html"/>
    <link rel="top" title="PaddlePaddle  documentation" href="../../index.html"/> 

  <link rel="stylesheet" href="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/css/perfect-scrollbar.min.css" type="text/css" />
  <link rel="stylesheet" href="../../_static/css/override.css" type="text/css" />
  <script>
  var _hmt = _hmt || [];
  (function() {
    var hm = document.createElement("script");
    hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
    var s = document.getElementsByTagName("script")[0]; 
    s.parentNode.insertBefore(hm, s);
  })();
  </script>

  

  
  <script src="../../_static/js/modernizr.min.js"></script>

</head>

<body class="wy-body-for-nav" role="document">

  
  <header class="site-header">
    <div class="site-logo">
      <a href="/"><img src="../../_static/images/PP_w.png"></a>
    </div>
    <div class="site-nav-links">
      <div class="site-menu">
        <a class="fork-on-github" href="https://github.com/PaddlePaddle/Paddle" target="_blank"><i class="fa fa-github"></i>Fork me on Github</a>
        <div class="language-switcher dropdown">
          <a type="button" data-toggle="dropdown">
            <span>English</span>
            <i class="fa fa-angle-up"></i>
            <i class="fa fa-angle-down"></i>
          </a>
          <ul class="dropdown-menu">
            <li><a href="/doc_cn">中文</a></li>
            <li><a href="/doc">English</a></li>
          </ul>
        </div>
        <ul class="site-page-links">
          <li><a href="/">Home</a></li>
        </ul>
      </div>
      <div class="doc-module">
        
        <ul>
<li class="toctree-l1"><a class="reference internal" href="../../getstarted/index_en.html">GET STARTED</a></li>
<li class="toctree-l1"><a class="reference internal" href="../index_en.html">HOW TO</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../api/index_en.html">API</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../mobile/index_en.html">MOBILE</a></li>
</ul>

        
<div role="search">
  <form id="rtd-search-form" class="wy-form" action="../../search.html" method="get">
    <input type="text" name="q" placeholder="Search docs" />
    <input type="hidden" name="check_keywords" value="yes" />
    <input type="hidden" name="area" value="default" />
  </form>
</div>        
      </div>
    </div>
  </header>
  
  <div class="main-content-wrap">

    
    <nav class="doc-menu-vertical" role="navigation">
        
          
          <ul>
<li class="toctree-l1"><a class="reference internal" href="../../getstarted/index_en.html">GET STARTED</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../../getstarted/build_and_install/index_en.html">Install and Build</a><ul>
T
Deploy to GitHub Pages: 0c0ff2828ccedb51db23290d6df9e4c83839d6af  
Travis CI 已提交
111 112 113 
<li class="toctree-l3"><a class="reference internal" href="../../getstarted/build_and_install/pip_install_en.html">Install Using pip</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../getstarted/build_and_install/docker_install_en.html">Run in Docker Containers</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../getstarted/build_and_install/build_from_source_en.html">Build from Sources</a></li>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 
</ul>
</li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../index_en.html">HOW TO</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../usage/cmd_parameter/index_en.html">Set Command-line Parameters</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../usage/cmd_parameter/use_case_en.html">Use Case</a></li>
<li class="toctree-l3"><a class="reference internal" href="../usage/cmd_parameter/arguments_en.html">Argument Outline</a></li>
<li class="toctree-l3"><a class="reference internal" href="../usage/cmd_parameter/detail_introduction_en.html">Detail Description</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../usage/cluster/cluster_train_en.html">PaddlePaddle Distributed Training</a></li>
<li class="toctree-l2"><a class="reference internal" href="../usage/k8s/k8s_en.html">Paddle On Kubernetes</a></li>
<li class="toctree-l2"><a class="reference internal" href="../usage/k8s/k8s_aws_en.html">Distributed PaddlePaddle Training on AWS with Kubernetes</a></li>
<li class="toctree-l2"><a class="reference internal" href="../dev/new_layer_en.html">Write New Layers</a></li>
<li class="toctree-l2"><a class="reference internal" href="../dev/contribute_to_paddle_en.html">Contribute Code</a></li>
<li class="toctree-l2"><a class="reference internal" href="../deep_model/rnn/index_en.html">RNN Models</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../deep_model/rnn/rnn_config_en.html">RNN Configuration</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="gpu_profiling_en.html">Tune GPU Performance</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../api/index_en.html">API</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../../api/v2/model_configs.html">Model Configuration</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/activation.html">Activation</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/layer.html">Layers</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/evaluators.html">Evaluators</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/optimizer.html">Optimizer</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/pooling.html">Pooling</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/networks.html">Networks</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/config/attr.html">Parameter Attribute</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../../api/v2/data.html">Data Reader Interface and DataSets</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/data/data_reader.html">Data Reader Interface</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/data/image.html">Image Interface</a></li>
<li class="toctree-l3"><a class="reference internal" href="../../api/v2/data/dataset.html">Dataset</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../../api/v2/run_logic.html">Training and Inference</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../mobile/index_en.html">MOBILE</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../../mobile/cross_compiling_for_android_en.html">Build PaddlePaddle for Android</a></li>
<li class="toctree-l2"><a class="reference internal" href="../../mobile/cross_compiling_for_raspberry_en.html">Build PaddlePaddle for Raspberry Pi</a></li>
</ul>
</li>
</ul>

        
    </nav>
    
    <section class="doc-content-wrap">

      

 







<div role="navigation" aria-label="breadcrumbs navigation">
  <ul class="wy-breadcrumbs">
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
182 
    <li>Profiling the Python Code</li>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
183 184 185 186 187 188 189 190 
  </ul>
</div>
      
      <div class="wy-nav-content" id="doc-content">
        <div class="rst-content">
          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
           <div itemprop="articleBody">
T
Deploy to GitHub Pages: 6dc5b34e5b9fc462f6cbd5c96883981fe8258264  
Travis CI 已提交
191 
  <p>This tutorial introduces techniques we use to profile and tune the
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
192 
CPU performance of PaddlePaddle.  We will use Python packages
T
Deploy to GitHub Pages: 6dc5b34e5b9fc462f6cbd5c96883981fe8258264  
Travis CI 已提交
193 194 
<code class="docutils literal"><span class="pre">cProfile</span></code> and <code class="docutils literal"><span class="pre">yep</span></code>, and Google&#8217;s <code class="docutils literal"><span class="pre">perftools</span></code>.</p>
<p>Profiling is the process that reveals performance bottlenecks,
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
195 
which could be very different from what&#8217;s in the developers&#8217; mind.
T
Deploy to GitHub Pages: 6dc5b34e5b9fc462f6cbd5c96883981fe8258264  
Travis CI 已提交
196 
Performance tuning is done to fix these bottlenecks. Performance optimization
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
197 
repeats the steps of profiling and tuning alternatively.</p>
T
Deploy to GitHub Pages: 6dc5b34e5b9fc462f6cbd5c96883981fe8258264  
Travis CI 已提交
198 
<p>PaddlePaddle users program AI applications by calling the Python API, which calls
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
199 200 201 202 203 204 205 206 207 208 209 210 211 
into <code class="docutils literal"><span class="pre">libpaddle.so.</span></code> written in C++.  In this tutorial, we focus on
the profiling and tuning of</p>
<ol class="simple">
<li>the Python code and</li>
<li>the mixture of Python and C++ code.</li>
</ol>
<div class="section" id="profiling-the-python-code">
<span id="profiling-the-python-code"></span><h1>Profiling the Python Code<a class="headerlink" href="#profiling-the-python-code" title="Permalink to this headline"></a></h1>
<div class="section" id="generate-the-performance-profiling-file">
<span id="generate-the-performance-profiling-file"></span><h2>Generate the Performance Profiling File<a class="headerlink" href="#generate-the-performance-profiling-file" title="Permalink to this headline"></a></h2>
<p>We can use Python standard
package, <a class="reference external" href="https://docs.python.org/2/library/profile.html"><code class="docutils literal"><span class="pre">cProfile</span></code></a>,
to generate Python profiling file.  For example:</p>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
212 213 214 
<div class="highlight-bash"><div class="highlight"><pre><span></span>python -m cProfile -o profile.out main.py
</pre></div>
</div>
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
215 216 217 
<p>where <code class="docutils literal"><span class="pre">main.py</span></code> is the program we are going to profile, <code class="docutils literal"><span class="pre">-o</span></code> specifies
the output file.  Without <code class="docutils literal"><span class="pre">-o</span></code>, <code class="docutils literal"><span class="pre">cProfile</span></code> would outputs to standard
output.</p>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
218 
</div>
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
219 220 221 222 223 
<div class="section" id="look-into-the-profiling-file">
<span id="look-into-the-profiling-file"></span><h2>Look into the Profiling File<a class="headerlink" href="#look-into-the-profiling-file" title="Permalink to this headline"></a></h2>
<p><code class="docutils literal"><span class="pre">cProfile</span></code> generates <code class="docutils literal"><span class="pre">profile.out</span></code> after <code class="docutils literal"><span class="pre">main.py</span></code> completes. We can
use <a class="reference external" href="https://github.com/ymichael/cprofilev"><code class="docutils literal"><span class="pre">cprofilev</span></code></a> to look into
the details:</p>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
224 225 226 
<div class="highlight-bash"><div class="highlight"><pre><span></span>cprofilev -a <span class="m">0</span>.0.0.0 -p <span class="m">3214</span> -f profile.out main.py
</pre></div>
</div>
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
227 228 229 230 231 232 233 234 235 
<p>where <code class="docutils literal"><span class="pre">-a</span></code> specifies the HTTP IP, <code class="docutils literal"><span class="pre">-p</span></code> specifies the port, <code class="docutils literal"><span class="pre">-f</span></code>
specifies the profiling file, and <code class="docutils literal"><span class="pre">main.py</span></code> is the source file.</p>
<p>Open the Web browser and points to the local IP and the specifies
port, we will see the output like the following:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span>   <span class="n">ncalls</span>  <span class="n">tottime</span>  <span class="n">percall</span>  <span class="n">cumtime</span>  <span class="n">percall</span> <span class="n">filename</span><span class="p">:</span><span class="n">lineno</span><span class="p">(</span><span class="n">function</span><span class="p">)</span>
        <span class="mi">1</span>    <span class="mf">0.284</span>    <span class="mf">0.284</span>   <span class="mf">29.514</span>   <span class="mf">29.514</span> <span class="n">main</span><span class="o">.</span><span class="n">py</span><span class="p">:</span><span class="mi">1</span><span class="p">(</span><span class="o">&lt;</span><span class="n">module</span><span class="o">&gt;</span><span class="p">)</span>
     <span class="mi">4696</span>    <span class="mf">0.128</span>    <span class="mf">0.000</span>   <span class="mf">15.748</span>    <span class="mf">0.003</span> <span class="o">/</span><span class="n">home</span><span class="o">/</span><span class="n">yuyang</span><span class="o">/</span><span class="n">perf_test</span><span class="o">/.</span><span class="n">env</span><span class="o">/</span><span class="n">lib</span><span class="o">/</span><span class="n">python2</span><span class="o">.</span><span class="mi">7</span><span class="o">/</span><span class="n">site</span><span class="o">-</span><span class="n">packages</span><span class="o">/</span><span class="n">paddle</span><span class="o">/</span><span class="n">v2</span><span class="o">/</span><span class="n">fluid</span><span class="o">/</span><span class="n">executor</span><span class="o">.</span><span class="n">py</span><span class="p">:</span><span class="mi">20</span><span class="p">(</span><span class="n">run</span><span class="p">)</span>
     <span class="mi">4696</span>   <span class="mf">12.040</span>    <span class="mf">0.003</span>   <span class="mf">12.040</span>    <span class="mf">0.003</span> <span class="p">{</span><span class="n">built</span><span class="o">-</span><span class="ow">in</span> <span class="n">method</span> <span class="n">run</span><span class="p">}</span>
        <span class="mi">1</span>    <span class="mf">0.144</span>    <span class="mf">0.144</span>    <span class="mf">6.534</span>    <span class="mf">6.534</span> <span class="o">/</span><span class="n">home</span><span class="o">/</span><span class="n">yuyang</span><span class="o">/</span><span class="n">perf_test</span><span class="o">/.</span><span class="n">env</span><span class="o">/</span><span class="n">lib</span><span class="o">/</span><span class="n">python2</span><span class="o">.</span><span class="mi">7</span><span class="o">/</span><span class="n">site</span><span class="o">-</span><span class="n">packages</span><span class="o">/</span><span class="n">paddle</span><span class="o">/</span><span class="n">v2</span><span class="o">/</span><span class="fm">__init__</span><span class="o">.</span><span class="n">py</span><span class="p">:</span><span class="mi">14</span><span class="p">(</span><span class="o">&lt;</span><span class="n">module</span><span class="o">&gt;</span><span class="p">)</span>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
236 237 
</pre></div>
</div>
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
238 239 240 
<p>where each line corresponds to Python function, and the meaning of
each column is as follows:</p>
<p>| column | meaning |
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
241 
| &#8212; | &#8212; |
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
242 243 244 245 246 247 248 
| ncalls | the number of calls into a function |
| tottime | the total execution time of the function, not including the
execution time of other functions called by the function |
| percall | tottime divided by ncalls |
| cumtime | the total execution time of the function, including the execution time of other functions being called |
| percall | cumtime divided by ncalls |
| filename:lineno(function) | where the function is defined |</p>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
249 
</div>
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
250 251 252 253 
<div class="section" id="identify-performance-bottlenecks">
<span id="identify-performance-bottlenecks"></span><h2>Identify Performance Bottlenecks<a class="headerlink" href="#identify-performance-bottlenecks" title="Permalink to this headline"></a></h2>
<p>Usually, <code class="docutils literal"><span class="pre">tottime</span></code> and the related <code class="docutils literal"><span class="pre">percall</span></code> time is what we want to
focus on. We can sort above profiling file by tottime:</p>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
254 255 256 257 258 259 260 
<div class="highlight-text"><div class="highlight"><pre><span></span>     4696   12.040    0.003   12.040    0.003 {built-in method run}
   300005    0.874    0.000    1.681    0.000 /home/yuyang/perf_test/.env/lib/python2.7/site-packages/paddle/v2/dataset/mnist.py:38(reader)
   107991    0.676    0.000    1.519    0.000 /home/yuyang/perf_test/.env/lib/python2.7/site-packages/paddle/v2/fluid/framework.py:219(__init__)
     4697    0.626    0.000    2.291    0.000 /home/yuyang/perf_test/.env/lib/python2.7/site-packages/paddle/v2/fluid/framework.py:428(sync_with_cpp)
        1    0.618    0.618    0.618    0.618 /home/yuyang/perf_test/.env/lib/python2.7/site-packages/paddle/v2/fluid/__init__.py:1(&lt;module&gt;)
</pre></div>
</div>
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
261 
<p>We can see that the most time-consuming function is the <code class="docutils literal"><span class="pre">built-in</span> <span class="pre">method</span> <span class="pre">run</span></code>, which is a C++ function in <code class="docutils literal"><span class="pre">libpaddle.so</span></code>.  We will
T
Deploy to GitHub Pages: 6dc5b34e5b9fc462f6cbd5c96883981fe8258264  
Travis CI 已提交
262 
explain how to profile C++ code in the next section.  At this
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
263 264 265 
moment, let&#8217;s look into the third function <code class="docutils literal"><span class="pre">sync_with_cpp</span></code>, which is a
Python function.  We can click it to understand more about it:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">Called</span> <span class="n">By</span><span class="p">:</span>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
266
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
267 268 
   <span class="n">Ordered</span> <span class="n">by</span><span class="p">:</span> <span class="n">internal</span> <span class="n">time</span>
   <span class="n">List</span> <span class="n">reduced</span> <span class="kn">from</span> <span class="mi">4497</span> <span class="n">to</span> <span class="mi">2</span> <span class="n">due</span> <span class="n">to</span> <span class="n">restriction</span> <span class="o">&lt;</span><span class="s1">&#39;sync_with_cpp&#39;</span><span class="o">&gt;</span>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
269
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
270 271 272 273 274 
<span class="n">Function</span>                                                                                                 <span class="n">was</span> <span class="n">called</span> <span class="n">by</span><span class="o">...</span>
                                                                                                             <span class="n">ncalls</span>  <span class="n">tottime</span>  <span class="n">cumtime</span>
<span class="o">/</span><span class="n">home</span><span class="o">/</span><span class="n">yuyang</span><span class="o">/</span><span class="n">perf_test</span><span class="o">/.</span><span class="n">env</span><span class="o">/</span><span class="n">lib</span><span class="o">/</span><span class="n">python2</span><span class="o">.</span><span class="mi">7</span><span class="o">/</span><span class="n">site</span><span class="o">-</span><span class="n">packages</span><span class="o">/</span><span class="n">paddle</span><span class="o">/</span><span class="n">v2</span><span class="o">/</span><span class="n">fluid</span><span class="o">/</span><span class="n">framework</span><span class="o">.</span><span class="n">py</span><span class="p">:</span><span class="mi">428</span><span class="p">(</span><span class="n">sync_with_cpp</span><span class="p">)</span>  <span class="o">&lt;-</span>    <span class="mi">4697</span>    <span class="mf">0.626</span>    <span class="mf">2.291</span>  <span class="o">/</span><span class="n">home</span><span class="o">/</span><span class="n">yuyang</span><span class="o">/</span><span class="n">perf_test</span><span class="o">/.</span><span class="n">env</span><span class="o">/</span><span class="n">lib</span><span class="o">/</span><span class="n">python2</span><span class="o">.</span><span class="mi">7</span><span class="o">/</span><span class="n">site</span><span class="o">-</span><span class="n">packages</span><span class="o">/</span><span class="n">paddle</span><span class="o">/</span><span class="n">v2</span><span class="o">/</span><span class="n">fluid</span><span class="o">/</span><span class="n">framework</span><span class="o">.</span><span class="n">py</span><span class="p">:</span><span class="mi">562</span><span class="p">(</span><span class="n">sync_with_cpp</span><span class="p">)</span>
<span class="o">/</span><span class="n">home</span><span class="o">/</span><span class="n">yuyang</span><span class="o">/</span><span class="n">perf_test</span><span class="o">/.</span><span class="n">env</span><span class="o">/</span><span class="n">lib</span><span class="o">/</span><span class="n">python2</span><span class="o">.</span><span class="mi">7</span><span class="o">/</span><span class="n">site</span><span class="o">-</span><span class="n">packages</span><span class="o">/</span><span class="n">paddle</span><span class="o">/</span><span class="n">v2</span><span class="o">/</span><span class="n">fluid</span><span class="o">/</span><span class="n">framework</span><span class="o">.</span><span class="n">py</span><span class="p">:</span><span class="mi">562</span><span class="p">(</span><span class="n">sync_with_cpp</span><span class="p">)</span>  <span class="o">&lt;-</span>    <span class="mi">4696</span>    <span class="mf">0.019</span>    <span class="mf">2.316</span>  <span class="o">/</span><span class="n">home</span><span class="o">/</span><span class="n">yuyang</span><span class="o">/</span><span class="n">perf_test</span><span class="o">/.</span><span class="n">env</span><span class="o">/</span><span class="n">lib</span><span class="o">/</span><span class="n">python2</span><span class="o">.</span><span class="mi">7</span><span class="o">/</span><span class="n">site</span><span class="o">-</span><span class="n">packages</span><span class="o">/</span><span class="n">paddle</span><span class="o">/</span><span class="n">v2</span><span class="o">/</span><span class="n">fluid</span><span class="o">/</span><span class="n">framework</span><span class="o">.</span><span class="n">py</span><span class="p">:</span><span class="mi">487</span><span class="p">(</span><span class="n">clone</span><span class="p">)</span>
                                                                                                                  <span class="mi">1</span>    <span class="mf">0.000</span>    <span class="mf">0.001</span>  <span class="o">/</span><span class="n">home</span><span class="o">/</span><span class="n">yuyang</span><span class="o">/</span><span class="n">perf_test</span><span class="o">/.</span><span class="n">env</span><span class="o">/</span><span class="n">lib</span><span class="o">/</span><span class="n">python2</span><span class="o">.</span><span class="mi">7</span><span class="o">/</span><span class="n">site</span><span class="o">-</span><span class="n">packages</span><span class="o">/</span><span class="n">paddle</span><span class="o">/</span><span class="n">v2</span><span class="o">/</span><span class="n">fluid</span><span class="o">/</span><span class="n">framework</span><span class="o">.</span><span class="n">py</span><span class="p">:</span><span class="mi">534</span><span class="p">(</span><span class="n">append_backward</span><span class="p">)</span>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
275 276
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
277 
<span class="n">Called</span><span class="p">:</span>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
278
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
279 280 
   <span class="n">Ordered</span> <span class="n">by</span><span class="p">:</span> <span class="n">internal</span> <span class="n">time</span>
   <span class="n">List</span> <span class="n">reduced</span> <span class="kn">from</span> <span class="mi">4497</span> <span class="n">to</span> <span class="mi">2</span> <span class="n">due</span> <span class="n">to</span> <span class="n">restriction</span> <span class="o">&lt;</span><span class="s1">&#39;sync_with_cpp&#39;</span><span class="o">&gt;</span>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
281 282 
</pre></div>
</div>
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
283 284 
<p>The lists of the callers of <code class="docutils literal"><span class="pre">sync_with_cpp</span></code> might help us understand
how to improve the function definition.</p>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
285 286 
</div>
</div>
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
287 288 289 290 291 292 293 294 295 296 297 
<div class="section" id="profiling-python-and-c-code">
<span id="profiling-python-and-c-code"></span><h1>Profiling Python and C++ Code<a class="headerlink" href="#profiling-python-and-c-code" title="Permalink to this headline"></a></h1>
<div class="section" id="generate-the-profiling-file">
<span id="generate-the-profiling-file"></span><h2>Generate the Profiling File<a class="headerlink" href="#generate-the-profiling-file" title="Permalink to this headline"></a></h2>
<p>To profile a mixture of Python and C++ code, we can use a Python
package, <code class="docutils literal"><span class="pre">yep</span></code>, that can work with Google&#8217;s <code class="docutils literal"><span class="pre">perftools</span></code>, which is a
commonly-used profiler for C/C++ code.</p>
<p>In Ubuntu systems, we can install <code class="docutils literal"><span class="pre">yep</span></code> and <code class="docutils literal"><span class="pre">perftools</span></code> by running the
following commands:</p>
<div class="highlight-bash"><div class="highlight"><pre><span></span>apt update
apt install libgoogle-perftools-dev
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
298 299 300 
pip install yep
</pre></div>
</div>
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
301 
<p>Then we can run the following command</p>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
302 303 304 
<div class="highlight-bash"><div class="highlight"><pre><span></span>python -m yep -v main.py
</pre></div>
</div>
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
305 306 307 
<p>to generate the profiling file.  The default filename is
<code class="docutils literal"><span class="pre">main.py.prof</span></code>.</p>
<p>Please be aware of the <code class="docutils literal"><span class="pre">-v</span></code> command line option, which prints the
T
Deploy to GitHub Pages: 6dc5b34e5b9fc462f6cbd5c96883981fe8258264  
Travis CI 已提交
308 309 
analysis results after generating the profiling file.  By examining the
the print result, we&#8217;d know that if we stripped debug
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
310 311 
information from <code class="docutils literal"><span class="pre">libpaddle.so</span></code> at build time.  The following hints
help make sure that the analysis results are readable:</p>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
312 
<ol class="simple">
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
313 314 315 316 317 318 319 320 321 322 323 324 
<li>Use GCC command line option <code class="docutils literal"><span class="pre">-g</span></code> when building <code class="docutils literal"><span class="pre">libpaddle.so</span></code> so to
include the debug information.  The standard building system of
PaddlePaddle is CMake, so you might want to set
<code class="docutils literal"><span class="pre">CMAKE_BUILD_TYPE=RelWithDebInfo</span></code>.</li>
<li>Use GCC command line option <code class="docutils literal"><span class="pre">-O2</span></code> or <code class="docutils literal"><span class="pre">-O3</span></code> to generate optimized
binary code. It doesn&#8217;t make sense to profile <code class="docutils literal"><span class="pre">libpaddle.so</span></code>
without optimization, because it would anyway run slowly.</li>
<li>Profiling the single-threaded binary file before the
multi-threading version, because the latter often generates tangled
profiling analysis result.  You might want to set environment
variable <code class="docutils literal"><span class="pre">OMP_NUM_THREADS=1</span></code> to prevents OpenMP from automatically
starting multiple threads.</li>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
325 326 
</ol>
</div>
T
Deploy to GitHub Pages: 6dc5b34e5b9fc462f6cbd5c96883981fe8258264  
Travis CI 已提交
327 328 329 
<div class="section" id="examining-the-profiling-file">
<span id="examining-the-profiling-file"></span><h2>Examining the Profiling File<a class="headerlink" href="#examining-the-profiling-file" title="Permalink to this headline"></a></h2>
<p>The tool we used to examine the profiling file generated by
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
330 331 332 333 
<code class="docutils literal"><span class="pre">perftools</span></code> is <a class="reference external" href="https://github.com/google/pprof"><code class="docutils literal"><span class="pre">pprof</span></code></a>, which
provides a Web-based GUI like <code class="docutils literal"><span class="pre">cprofilev</span></code>.</p>
<p>We can rely on the standard Go toolchain to retrieve the source code
of <code class="docutils literal"><span class="pre">pprof</span></code> and build it:</p>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
334 335 336 
<div class="highlight-bash"><div class="highlight"><pre><span></span>go get github.com/google/pprof
</pre></div>
</div>
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
337 338 
<p>Then we can use it to profile <code class="docutils literal"><span class="pre">main.py.prof</span></code> generated in the previous
section:</p>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
339 340 341 
<div class="highlight-bash"><div class="highlight"><pre><span></span>pprof -http<span class="o">=</span><span class="m">0</span>.0.0.0:3213 <span class="sb">`</span>which python<span class="sb">`</span>  ./main.py.prof
</pre></div>
</div>
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
342 343 344 
<p>Where <code class="docutils literal"><span class="pre">-http</span></code> specifies the IP and port of the HTTP service.
Directing our Web browser to the service, we would see something like
the following:</p>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
345 346 
<p><img alt="result" src="../../_images/pprof_1.png" /></p>
</div>
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
347 348 349 350 
<div class="section" id="identifying-the-performance-bottlenecks">
<span id="identifying-the-performance-bottlenecks"></span><h2>Identifying the Performance Bottlenecks<a class="headerlink" href="#identifying-the-performance-bottlenecks" title="Permalink to this headline"></a></h2>
<p>Similar to how we work with <code class="docutils literal"><span class="pre">cprofilev</span></code>, we&#8217;d focus on <code class="docutils literal"><span class="pre">tottime</span></code> and
<code class="docutils literal"><span class="pre">cumtime</span></code>.</p>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
351 
<p><img alt="kernel_perf" src="../../_images/pprof_2.png" /></p>
T
Deploy to GitHub Pages: 605b3e449911420e5a171085d457916d668268e1  
Travis CI 已提交
352 353 354 355 356 
<p>We can see that the execution time of multiplication and the computing
of the gradient of multiplication takes 2% to 4% of the total running
time, and <code class="docutils literal"><span class="pre">MomentumOp</span></code> takes about 17%. Obviously, we&#8217;d want to
optimize <code class="docutils literal"><span class="pre">MomentumOp</span></code>.</p>
<p><code class="docutils literal"><span class="pre">pprof</span></code> would mark performance critical parts of the program in
T
Deploy to GitHub Pages: 6dc5b34e5b9fc462f6cbd5c96883981fe8258264  
Travis CI 已提交
357 
red. It&#8217;s a good idea to follow the hints.</p>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 
</div>
</div>


           </div>
          </div>
          <footer>
  

  <hr/>

  <div role="contentinfo">
    <p>
        &copy; Copyright 2016, PaddlePaddle developers.

    </p>
  </div>
  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 

</footer>

        </div>
      </div>

    </section>

  </div>
  


  

    <script type="text/javascript">
        var DOCUMENTATION_OPTIONS = {
            URL_ROOT:'../../',
            VERSION:'',
            COLLAPSE_INDEX:false,
            FILE_SUFFIX:'.html',
            HAS_SOURCE:  true,
            SOURCELINK_SUFFIX: ".txt",
        };
    </script>
      <script type="text/javascript" src="../../_static/jquery.js"></script>
      <script type="text/javascript" src="../../_static/underscore.js"></script>
      <script type="text/javascript" src="../../_static/doctools.js"></script>
      <script type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
       
  

  
  
    <script type="text/javascript" src="../../_static/js/theme.js"></script>
  
  
  <script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.7/js/bootstrap.min.js" integrity="sha384-Tc5IQib027qvyjSMfHjOMaLkfuWVxZxUPnCJA7l2mCWNIpG9mGCD8wGNIcPD7Txa" crossorigin="anonymous"></script>
  <script src="https://cdn.jsdelivr.net/perfect-scrollbar/0.6.14/js/perfect-scrollbar.jquery.min.js"></script>
  <script src="../../_static/js/paddle_doc_init.js"></script> 

</body>
</html>