graph.html



<!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>Design Doc: Computations as a Graph &mdash; PaddlePaddle  文档</title>
  

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

        <link rel="index" title="索引"
              href="../genindex.html"/>
        <link rel="search" title="搜索" href="../search.html"/>
    <link rel="top" title="PaddlePaddle  文档" href="../index.html"/> 
<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">

  <div class="wy-grid-for-nav">

    
    <nav data-toggle="wy-nav-shift" class="wy-nav-side">
      <div class="wy-side-scroll">
        <div class="wy-side-nav-search">
          

            <a href="../index_cn.html" class="icon icon-home"> PaddlePaddle
          

          </a>

          
<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 class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation">
          
<nav class="doc-menu-vertical" role="navigation">

<ul>
<li class="toctree-l1"><a class="reference internal" href="../getstarted/index_cn.html">新手入门</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../getstarted/quickstart_cn.html">快速开始</a></li>
<li class="toctree-l2"><a class="reference internal" href="../getstarted/concepts/use_concepts_cn.html">基本使用概念</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../build_and_install/index_cn.html">安装与编译</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../build_and_install/pip_install_cn.html">使用pip安装</a></li>
<li class="toctree-l2"><a class="reference internal" href="../build_and_install/docker_install_cn.html">使用Docker安装运行</a></li>
<li class="toctree-l2"><a class="reference internal" href="../build_and_install/build_from_source_cn.html">从源码编译</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../howto/index_cn.html">进阶使用</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../howto/cmd_parameter/index_cn.html">命令行参数设置</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../howto/cmd_parameter/use_case_cn.html">使用案例</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cmd_parameter/arguments_cn.html">参数概述</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cmd_parameter/detail_introduction_cn.html">细节描述</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../howto/cluster/index_cn.html">分布式训练</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../howto/cluster/preparations_cn.html">环境准备</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cluster/cmd_argument_cn.html">启动参数说明</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cluster/multi_cluster/index_cn.html">在不同集群中运行</a><ul>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/k8s_cn.html">Kubernetes单机训练</a></li>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/k8s_distributed_cn.html">Kubernetes分布式训练</a></li>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/openmpi_cn.html">在OpenMPI集群中启动训练</a></li>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/fabric_cn.html">使用fabric启动集群训练</a></li>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/k8s_aws_cn.html">Kubernetes on AWS</a></li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../howto/capi/index_cn.html">C-API预测库</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../howto/capi/compile_paddle_lib_cn.html">安装与编译C-API预测库</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/capi/organization_of_the_inputs_cn.html">输入/输出数据组织</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/capi/workflow_of_capi_cn.html">C-API使用流程</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../howto/rnn/index_cn.html">RNN模型</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/rnn_config_cn.html">RNN配置</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/recurrent_group_cn.html">Recurrent Group教程</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/hierarchical_layer_cn.html">支持双层序列作为输入的Layer</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/hrnn_rnn_api_compare_cn.html">单双层RNN API对比介绍</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../howto/optimization/gpu_profiling_cn.html">GPU性能调优</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../dev/index_cn.html">开发标准</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../dev/contribute_to_paddle_cn.html">如何贡献代码</a></li>
<li class="toctree-l2"><a class="reference internal" href="../dev/write_docs_cn.html">如何贡献文档</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../faq/index_cn.html">FAQ</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../faq/build_and_install/index_cn.html">编译安装与单元测试</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/model/index_cn.html">模型配置</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/parameter/index_cn.html">参数设置</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/local/index_cn.html">本地训练与预测</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/cluster/index_cn.html">集群训练与预测</a></li>
</ul>
</li>
</ul>

</nav>

        </div>
      </div>
    </nav>

    <section data-toggle="wy-nav-shift" class="wy-nav-content-wrap">

      
      <nav class="wy-nav-top" role="navigation" aria-label="top navigation">
        <i data-toggle="wy-nav-top" class="fa fa-bars"></i>
        <a href="../index_cn.html">PaddlePaddle</a>
      </nav>


      <div class="wy-nav-content">
        <div class="rst-content">
          

<div role="navigation" aria-label="breadcrumbs navigation">
  <ul class="wy-breadcrumbs">
    <li><a href="../index_cn.html">Docs</a> &raquo;</li>
      
    <li>Design Doc: Computations as a Graph</li>
      <li class="wy-breadcrumbs-aside">
        
          
            <a href="../_sources/design/graph.md.txt" rel="nofollow"> View page source</a>
          
        
      </li>
  </ul>
  <hr/>
</div>
          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
           <div itemprop="articleBody">
            
  <div class="section" id="design-doc-computations-as-a-graph">
<span id="design-doc-computations-as-a-graph"></span><h1>Design Doc: Computations as a Graph<a class="headerlink" href="#design-doc-computations-as-a-graph" title="永久链接至标题">¶</a></h1>
<p>A primary goal of the refactorization of PaddlePaddle is a more flexible representation of deep learning computation, in particular, a graph of operators and variables, instead of sequences of layers as before.</p>
<p>This document explains that the construction of a graph as three steps:</p>
<ul class="simple">
<li>construct the forward part</li>
<li>construct the backward part</li>
<li>construct the optimization part</li>
</ul>
<div class="section" id="the-construction-of-a-graph">
<span id="the-construction-of-a-graph"></span><h2>The Construction of a Graph<a class="headerlink" href="#the-construction-of-a-graph" title="永久链接至标题">¶</a></h2>
<p>Let us take the problem of image classification as a simple example.  The application program that trains the model looks like:</p>
<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">x</span> <span class="o">=</span> <span class="n">layer</span><span class="o">.</span><span class="n">data</span><span class="p">(</span><span class="s2">&quot;images&quot;</span><span class="p">)</span>
<span class="n">l</span> <span class="o">=</span> <span class="n">layer</span><span class="o">.</span><span class="n">data</span><span class="p">(</span><span class="s2">&quot;label&quot;</span><span class="p">)</span>
<span class="n">y</span> <span class="o">=</span> <span class="n">layer</span><span class="o">.</span><span class="n">fc</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="n">cost</span> <span class="o">=</span> <span class="n">layer</span><span class="o">.</span><span class="n">mse</span><span class="p">(</span><span class="n">y</span><span class="p">,</span> <span class="n">l</span><span class="p">)</span>
<span class="n">optimize</span><span class="p">(</span><span class="n">cost</span><span class="p">)</span>
<span class="n">train</span><span class="p">(</span><span class="n">cost</span><span class="p">,</span> <span class="n">reader</span><span class="o">=</span><span class="n">mnist</span><span class="o">.</span><span class="n">train</span><span class="p">())</span>
</pre></div>
</div>
<div class="section" id="forward-part">
<span id="forward-part"></span><h3>Forward Part<a class="headerlink" href="#forward-part" title="永久链接至标题">¶</a></h3>
<p>The first four lines of above program build the forward part of the graph.</p>
<p><img alt="" src="../_images/graph_construction_example_forward_only.png" /></p>
<p>In particular, the first line <code class="docutils literal"><span class="pre">x</span> <span class="pre">=</span> <span class="pre">layer.data(&quot;images&quot;)</span></code> creates variable x and a Feed operator that copies a column from the minibatch to x.  <code class="docutils literal"><span class="pre">y</span> <span class="pre">=</span> <span class="pre">layer.fc(x)</span></code> creates not only the FC operator and output variable y, but also two parameters, W and b, and the initialization operators.</p>
<p>Initialization operators are kind of &#8220;run-once&#8221; operators &#8211; the <code class="docutils literal"><span class="pre">Run</span></code> method increments a class data member counter so to run at most once.  By doing so, a parameter wouldn&#8217;t be initialized repeatedly, say, in every minibatch.</p>
<p>In this example, all operators are created as <code class="docutils literal"><span class="pre">OpDesc</span></code> protobuf messages, and all variables are <code class="docutils literal"><span class="pre">VarDesc</span></code>.  These protobuf messages are saved in a <code class="docutils literal"><span class="pre">BlockDesc</span></code> protobuf message.</p>
</div>
<div class="section" id="backward-part">
<span id="backward-part"></span><h3>Backward Part<a class="headerlink" href="#backward-part" title="永久链接至标题">¶</a></h3>
<p>The fifth line <code class="docutils literal"><span class="pre">optimize(cost)</span></code> calls two functions, <code class="docutils literal"><span class="pre">ConstructBackwardGraph</span></code> and <code class="docutils literal"><span class="pre">ConstructOptimizationGraph</span></code>.</p>
<p><code class="docutils literal"><span class="pre">ConstructBackwardGraph</span></code> traverses the forward graph in the <code class="docutils literal"><span class="pre">BlockDesc</span></code> protobuf message and builds the backward part.</p>
<p><img alt="" src="../_images/graph_construction_example_forward_backward.png" /></p>
<p>According to the chain rule of gradient computation, <code class="docutils literal"><span class="pre">ConstructBackwardGraph</span></code> would</p>
<ol class="simple">
<li>create a gradient operator G for each operator F,</li>
<li>make all inputs, outputs, and outputs&#8217; gradient of F as inputs of G,</li>
<li>create gradients for all inputs of F, except for those who don&#8217;t have gradients, like x and l, and</li>
<li>make all these gradients as outputs of G.</li>
</ol>
</div>
<div class="section" id="optimization-part">
<span id="optimization-part"></span><h3>Optimization Part<a class="headerlink" href="#optimization-part" title="永久链接至标题">¶</a></h3>
<p>For each parameter, like W and b created by <code class="docutils literal"><span class="pre">layer.fc</span></code>, marked as double circles in above graphs, <code class="docutils literal"><span class="pre">ConstructOptimizationGraph</span></code> creates an optimization operator to apply its gradient.  Here results in the complete graph:</p>
<p><img alt="" src="../_images/graph_construction_example_all.png" /></p>
</div>
</div>
<div class="section" id="block-and-graph">
<span id="block-and-graph"></span><h2>Block and Graph<a class="headerlink" href="#block-and-graph" title="永久链接至标题">¶</a></h2>
<p>The word block and graph are interchangable in the desgin of PaddlePaddle.  A <a class="reference external" href="https://github.com/PaddlePaddle/Paddle/pull/3708">Block</a> is a metaphore of the code and local variables in a pair of curly braces in programming languages, where operators are like statements or instructions.  A graph of operators and variables is a representation of the block.</p>
<p>A Block keeps operators in an array <code class="docutils literal"><span class="pre">BlockDesc::ops</span></code></p>
<div class="highlight-protobuf"><div class="highlight"><pre><span></span><span class="kd">message</span> <span class="nc">BlockDesc</span> <span class="p">{</span>
  <span class="k">repeated</span> <span class="n">OpDesc</span> <span class="na">ops</span> <span class="o">=</span> <span class="mi">1</span><span class="p">;</span>
  <span class="k">repeated</span> <span class="n">VarDesc</span> <span class="na">vars</span> <span class="o">=</span> <span class="mi">2</span><span class="p">;</span>
<span class="p">}</span>
</pre></div>
</div>
<p>in the order that they appear in user programs, like the Python program at the beginning of this article.  We can imagine that in <code class="docutils literal"><span class="pre">ops</span></code>,  we have some forward operators, followed by some gradient operators, and then some optimization operators.</p>
</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
        };
    </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="../_static/translations.js"></script>
      <script type="text/javascript" src="https://cdn.bootcss.com/mathjax/2.7.0/MathJax.js"></script>

  
    <script type="text/javascript" src="../_static/js/theme.js"></script>
  

  <script type="text/javascript">
      jQuery(function () {
          SphinxRtdTheme.StickyNav.enable();
      });
  </script>
   

</body>
</html>