Skip to content

P
Paddle

PaddlePaddle / Paddle
大约 1 年 前同步成功

通知 2298
Star 20931
Fork 5422
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 24.3 KB
Newer Older
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
1 2 3 4 5 6 7 8 9 10 11 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 


<!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">
  
  <title>Python代码的性能分析 &mdash; PaddlePaddle  documentation</title>
  

  
  

  

  
  
    

  

  
  
    <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
114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 
</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">
      
    <li>Python代码的性能分析</li>
  </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">
            
  <p>此教程会介绍如何使用Python的cProfile包，与Python库yep，google perftools来运行性能分析(Profiling)与调优。</p>
<p>运行性能分析可以让开发人员科学的，有条不紊的对程序进行性能优化。性能分析是性能调优的基础。因为在程序实际运行中，真正的瓶颈可能和程序员开发过程中想象的瓶颈相去甚远。</p>
<p>性能优化的步骤，通常是循环重复若干次『性能分析 &#8211;&gt; 寻找瓶颈 &#8212;&gt; 调优瓶颈 &#8211;&gt; 性能分析确认调优效果』。其中性能分析是性能调优的至关重要的量化指标。</p>
<p>Paddle提供了Python语言绑定。用户使用Python进行神经网络编程，训练，测试。Python解释器通过<code class="docutils literal"><span class="pre">pybind</span></code>和<code class="docutils literal"><span class="pre">swig</span></code>调用Paddle的动态链接库，进而调用Paddle C++部分的代码。所以Paddle的性能分析与调优分为两个部分:</p>
<ul class="simple">
<li>Python代码的性能分析</li>
<li>Python与C++混合代码的性能分析</li>
</ul>
<div class="section" id="python">
<span id="python"></span><h1>Python代码的性能分析<a class="headerlink" href="#python" title="Permalink to this headline"></a></h1>
<div class="section" id="">
<span id="id1"></span><h2>生成性能分析文件<a class="headerlink" href="#" title="Permalink to this headline"></a></h2>
<p>Python标准库中提供了性能分析的工具包，<a class="reference external" href="https://docs.python.org/2/library/profile.html">cProfile</a>。生成Python性能分析的命令如下:</p>
<div class="highlight-bash"><div class="highlight"><pre><span></span>python -m cProfile -o profile.out main.py
</pre></div>
</div>
<p>其中<code class="docutils literal"><span class="pre">-o</span></code>标识了一个输出的文件名，用来存储本次性能分析的结果。如果不指定这个文件，<code class="docutils literal"><span class="pre">cProfile</span></code>会打印一些统计信息到<code class="docutils literal"><span class="pre">stdout</span></code>。这不方便我们进行后期处理(进行<code class="docutils literal"><span class="pre">sort</span></code>, <code class="docutils literal"><span class="pre">split</span></code>, <code class="docutils literal"><span class="pre">cut</span></code>等等)。</p>
</div>
<div class="section" id="">
<span id="id2"></span><h2>查看性能分析文件<a class="headerlink" href="#" title="Permalink to this headline"></a></h2>
<p>当main.py运行完毕后，性能分析结果文件<code class="docutils literal"><span class="pre">profile.out</span></code>就生成出来了。我们可以使用<a class="reference external" href="https://github.com/ymichael/cprofilev">cprofilev</a>来查看性能分析结果。<code class="docutils literal"><span class="pre">cprofilev</span></code>是一个Python的第三方库。使用它会开启一个HTTP服务，将性能分析结果以网页的形式展示出来。</p>
<p>使用<code class="docutils literal"><span class="pre">pip</span> <span class="pre">install</span> <span class="pre">cprofilev</span></code>安装<code class="docutils literal"><span class="pre">cprofilev</span></code>工具。安装完成后，使用如下命令开启HTTP服务</p>
<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>
<p>其中<code class="docutils literal"><span class="pre">-a</span></code>标识HTTP服务绑定的IP。使用<code class="docutils literal"><span class="pre">0.0.0.0</span></code>允许外网访问这个HTTP服务。<code class="docutils literal"><span class="pre">-p</span></code>标识HTTP服务的端口。<code class="docutils literal"><span class="pre">-f</span></code>标识性能分析的结果文件。<code class="docutils literal"><span class="pre">main.py</span></code>标识被性能分析的源文件。</p>
<p>访问对应网址，即可显示性能分析的结果。性能分析结果格式如下:</p>
<div class="highlight-text"><div class="highlight"><pre><span></span>   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
        1    0.284    0.284   29.514   29.514 main.py:1(&lt;module&gt;)
     4696    0.128    0.000   15.748    0.003 /home/yuyang/perf_test/.env/lib/python2.7/site-packages/paddle/v2/fluid/executor.py:20(run)
     4696   12.040    0.003   12.040    0.003 {built-in method run}
        1    0.144    0.144    6.534    6.534 /home/yuyang/perf_test/.env/lib/python2.7/site-packages/paddle/v2/__init__.py:14(&lt;module&gt;)
</pre></div>
</div>
<p>每一列的含义是:</p>
<p>| 列名 | 含义 |
| &#8212; | &#8212; |
| ncalls | 函数的调用次数 |
| tottime | 函数实际使用的总时间。该时间去除掉本函数调用其他函数的时间 |
| percall | tottime的每次调用平均时间 |
| cumtime | 函数总时间。包含这个函数调用其他函数的时间 |
| percall | cumtime的每次调用平均时间 |
| filename:lineno(function) | 文件名, 行号，函数名 |</p>
</div>
<div class="section" id="">
<span id="id3"></span><h2>寻找性能瓶颈<a class="headerlink" href="#" title="Permalink to this headline"></a></h2>
<p>通常<code class="docutils literal"><span class="pre">tottime</span></code>和<code class="docutils literal"><span class="pre">cumtime</span></code>是寻找瓶颈的关键指标。这两个指标代表了某一个函数真实的运行时间。</p>
<p>将性能分析结果按照tottime排序，效果如下:</p>
<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: 7300655ffd5deb47e24e493524534c94570ed48b  
Travis CI 已提交
246 
<p>可以看到最耗时的函数是C++端的<code class="docutils literal"><span class="pre">run</span></code>函数。这需要联合我们第二节<code class="docutils literal"><span class="pre">Python</span></code>与<code class="docutils literal"><span class="pre">C++</span></code>混合代码的性能分析来进行调优。而<code class="docutils literal"><span class="pre">sync_with_cpp</span></code>函数的总共耗时很长，每次调用的耗时也很长。于是我们可以点击<code class="docutils literal"><span class="pre">sync_with_cpp</span></code>的详细信息，了解其调用关系。</p>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 
<div class="highlight-text"><div class="highlight"><pre><span></span>Called By:

   Ordered by: internal time
   List reduced from 4497 to 2 due to restriction &lt;&#39;sync_with_cpp&#39;&gt;

Function                                                                                                 was called by...
                                                                                                             ncalls  tottime  cumtime
/home/yuyang/perf_test/.env/lib/python2.7/site-packages/paddle/v2/fluid/framework.py:428(sync_with_cpp)  &lt;-    4697    0.626    2.291  /home/yuyang/perf_test/.env/lib/python2.7/site-packages/paddle/v2/fluid/framework.py:562(sync_with_cpp)
/home/yuyang/perf_test/.env/lib/python2.7/site-packages/paddle/v2/fluid/framework.py:562(sync_with_cpp)  &lt;-    4696    0.019    2.316  /home/yuyang/perf_test/.env/lib/python2.7/site-packages/paddle/v2/fluid/framework.py:487(clone)
                                                                                                                  1    0.000    0.001  /home/yuyang/perf_test/.env/lib/python2.7/site-packages/paddle/v2/fluid/framework.py:534(append_backward)


Called:

   Ordered by: internal time
   List reduced from 4497 to 2 due to restriction &lt;&#39;sync_with_cpp&#39;&gt;
</pre></div>
</div>
<p>通常观察热点函数间的调用关系，和对应行的代码，就可以了解到问题代码在哪里。当我们做出性能修正后，再次进行性能分析(profiling)即可检查我们调优后的修正是否能够改善程序的性能。</p>
</div>
</div>
<div class="section" id="pythonc">
<span id="pythonc"></span><h1>Python与C++混合代码的性能分析<a class="headerlink" href="#pythonc" title="Permalink to this headline"></a></h1>
<div class="section" id="">
<span id="id4"></span><h2>生成性能分析文件<a class="headerlink" href="#" title="Permalink to this headline"></a></h2>
<p>C++的性能分析工具非常多。常见的包括<code class="docutils literal"><span class="pre">gprof</span></code>, <code class="docutils literal"><span class="pre">valgrind</span></code>, <code class="docutils literal"><span class="pre">google-perftools</span></code>。但是调试Python中使用的动态链接库与直接调试原始二进制相比增加了很多复杂度。幸而Python的一个第三方库<code class="docutils literal"><span class="pre">yep</span></code>提供了方便的和<code class="docutils literal"><span class="pre">google-perftools</span></code>交互的方法。于是这里使用<code class="docutils literal"><span class="pre">yep</span></code>进行Python与C++混合代码的性能分析</p>
<p>使用<code class="docutils literal"><span class="pre">yep</span></code>前需要安装<code class="docutils literal"><span class="pre">google-perftools</span></code>与<code class="docutils literal"><span class="pre">yep</span></code>包。ubuntu下安装命令为</p>
<div class="highlight-bash"><div class="highlight"><pre><span></span>apt install libgoogle-perftools-dev
pip install yep
</pre></div>
</div>
<p>安装完毕后，我们可以通过</p>
<div class="highlight-bash"><div class="highlight"><pre><span></span>python -m yep -v main.py
</pre></div>
</div>
<p>生成性能分析文件。生成的性能分析文件为<code class="docutils literal"><span class="pre">main.py.prof</span></code>。</p>
<p>命令行中的<code class="docutils literal"><span class="pre">-v</span></code>指定在生成性能分析文件之后，在命令行显示分析结果。我们可以在命令行中简单的看一下生成效果。因为C++与Python不同，编译时可能会去掉调试信息，运行时也可能因为多线程产生混乱不可读的性能分析结果。为了生成更可读的性能分析结果，可以采取下面几点措施:</p>
<ol class="simple">
<li>编译时指定<code class="docutils literal"><span class="pre">-g</span></code>生成调试信息。使用cmake的话，可以将CMAKE_BUILD_TYPE指定为<code class="docutils literal"><span class="pre">RelWithDebInfo</span></code>。</li>
<li>编译时一定要开启优化。单纯的<code class="docutils literal"><span class="pre">Debug</span></code>编译性能会和<code class="docutils literal"><span class="pre">-O2</span></code>或者<code class="docutils literal"><span class="pre">-O3</span></code>有非常大的差别。<code class="docutils literal"><span class="pre">Debug</span></code>模式下的性能测试是没有意义的。</li>
T
Deploy to GitHub Pages: 7300655ffd5deb47e24e493524534c94570ed48b  
Travis CI 已提交
287 
<li>运行性能分析的时候，先从单线程开始，再开启多线程，进而多机。毕竟单线程调试更容易。可以设置<code class="docutils literal"><span class="pre">OMP_NUM_THREADS=1</span></code>这个环境变量关闭openmp优化。</li>
T
Deploy to GitHub Pages: 7c57f90dc2e95704e71adc40931581d1be437f32  
Travis CI 已提交
288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 
</ol>
</div>
<div class="section" id="">
<span id="id5"></span><h2>查看性能分析文件<a class="headerlink" href="#" title="Permalink to this headline"></a></h2>
<p>在运行完性能分析后，会生成性能分析结果文件。我们可以使用<a class="reference external" href="https://github.com/google/pprof">pprof</a>来显示性能分析结果。注意，这里使用了用<code class="docutils literal"><span class="pre">Go</span></code>语言重构后的<code class="docutils literal"><span class="pre">pprof</span></code>，因为这个工具具有web服务界面，且展示效果更好。</p>
<p>安装<code class="docutils literal"><span class="pre">pprof</span></code>的命令和一般的<code class="docutils literal"><span class="pre">Go</span></code>程序是一样的，其命令如下:</p>
<div class="highlight-bash"><div class="highlight"><pre><span></span>go get github.com/google/pprof
</pre></div>
</div>
<p>进而我们可以使用如下命令开启一个HTTP服务:</p>
<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>
<p>这行命令中，<code class="docutils literal"><span class="pre">-http</span></code>指开启HTTP服务。<code class="docutils literal"><span class="pre">which</span> <span class="pre">python</span></code>会产生当前Python二进制的完整路径，进而指定了Python可执行文件的路径。<code class="docutils literal"><span class="pre">./main.py.prof</span></code>输入了性能分析结果。</p>
<p>访问对应的网址，我们可以查看性能分析的结果。结果如下图所示:</p>
<p><img alt="result" src="../../_images/pprof_1.png" /></p>
</div>
<div class="section" id="">
<span id="id6"></span><h2>寻找性能瓶颈<a class="headerlink" href="#" title="Permalink to this headline"></a></h2>
<p>与寻找Python代码的性能瓶颈类似，寻找Python与C++混合代码的性能瓶颈也是要看<code class="docutils literal"><span class="pre">tottime</span></code>和<code class="docutils literal"><span class="pre">cumtime</span></code>。而<code class="docutils literal"><span class="pre">pprof</span></code>展示的调用图也可以帮助我们发现性能中的问题。</p>
<p>例如下图中，</p>
<p><img alt="kernel_perf" src="../../_images/pprof_2.png" /></p>
<p>在一次训练中，乘法和乘法梯度的计算占用2%-4%左右的计算时间。而<code class="docutils literal"><span class="pre">MomentumOp</span></code>占用了17%左右的计算时间。显然，<code class="docutils literal"><span class="pre">MomentumOp</span></code>的性能有问题。</p>
<p>在<code class="docutils literal"><span class="pre">pprof</span></code>中，对于性能的关键路径都做出了红色标记。先检查关键路径的性能问题，再检查其他部分的性能问题，可以更有次序的完成性能的优化。</p>
</div>
</div>
<div class="section" id="">
<span id="id7"></span><h1>总结<a class="headerlink" href="#" title="Permalink to this headline"></a></h1>
<p>至此，两种性能分析的方式都介绍完毕了。希望通过这两种性能分析的方式，Paddle的开发人员和使用人员可以有次序的，科学的发现和解决性能问题。</p>
</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>