optimizer.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>Optimizer Design &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"/> 
<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_en.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_en.html">GET STARTED</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../getstarted/quickstart_en.html">Quick Start</a></li>
<li class="toctree-l2"><a class="reference internal" href="../getstarted/concepts/use_concepts_en.html">Basic Concept</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../build_and_install/index_en.html">Install and Build</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../build_and_install/pip_install_en.html">Install using pip</a></li>
<li class="toctree-l2"><a class="reference internal" href="../build_and_install/docker_install_en.html">Run in Docker Containers</a></li>
<li class="toctree-l2"><a class="reference internal" href="../build_and_install/build_from_source_en.html">Build from Sources</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../howto/index_en.html">HOW TO</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../howto/cmd_parameter/index_en.html">Set Command-line Parameters</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../howto/cmd_parameter/use_case_en.html">Use Case</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cmd_parameter/arguments_en.html">Argument Outline</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cmd_parameter/detail_introduction_en.html">Detail Description</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../howto/cluster/index_en.html">Distributed Training</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../howto/cluster/preparations_en.html">Preparations</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cluster/cmd_argument_en.html">Command-line arguments</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cluster/multi_cluster/index_en.html">Use different clusters</a><ul>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/fabric_en.html">Fabric</a></li>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/openmpi_en.html">OpenMPI</a></li>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/k8s_en.html">Kubernetes</a></li>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/k8s_aws_en.html">Kubernetes on AWS</a></li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../howto/rnn/index_en.html">RNN Models</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/rnn_config_en.html">RNN Configuration</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/recurrent_group_en.html">Recurrent Group Tutorial</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/hierarchical_layer_en.html">Layers supporting hierarchical sequence as input</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/hrnn_rnn_api_compare_en.html">API comparision between RNN and hierarchical RNN</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../howto/optimization/gpu_profiling_en.html">Tune GPU Performance</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../dev/index_en.html">Development</a><ul>
<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="../dev/write_docs_en.html">Contribute Documentation</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../faq/index_en.html">FAQ</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../faq/build_and_install/index_en.html">Install, Build and Unit test</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/model/index_en.html">Model Configuration</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/parameter/index_en.html">Parameter Setting</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/local/index_en.html">Local Training and Prediction</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/cluster/index_en.html">Cluster Training and Prediction</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_en.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_en.html">Docs</a> &raquo;</li>
      
    <li>Optimizer Design</li>
      <li class="wy-breadcrumbs-aside">
        
          
            <a href="../_sources/design/optimizer.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="optimizer-design">
<span id="optimizer-design"></span><h1>Optimizer Design<a class="headerlink" href="#optimizer-design" title="Permalink to this headline">¶</a></h1>
<div class="section" id="the-problem">
<span id="the-problem"></span><h2>The Problem<a class="headerlink" href="#the-problem" title="Permalink to this headline">¶</a></h2>
<p>A PaddlePaddle program, or a block, is a sequence of operators operating variables.  A training program needs to do three kinds of works:</p>
<ol class="simple">
<li>the forward pass, which computes intermediate results and the cost(s),</li>
<li>the backward pass, which derives gradients from intermediate results and costs, and</li>
<li>the optimization pass, which update model parameters to optimize the cost(s).</li>
</ol>
<p>These works rely on three kinds of operators:</p>
<ol class="simple">
<li>forward operators,</li>
<li>gradient operators, and</li>
<li>optimization operators.</li>
</ol>
<p>It&#8217;s true that users should be able to create all these operators manually by calling some low-level API, but it would be much more convenient if they could only describe the forward pass and let PaddlePaddle create the backward and optimization operators automatically.</p>
<p>In this design, we propose a high-level API that automatically derives the optimisation pass and operators from the forward pass.</p>
</div>
<div class="section" id="high-level-python-api-to-describe-the-training-process">
<span id="high-level-python-api-to-describe-the-training-process"></span><h2>High-level Python API to describe the training process<a class="headerlink" href="#high-level-python-api-to-describe-the-training-process" title="Permalink to this headline">¶</a></h2>
<ol>
<li><p class="first">User write code to describe the network:</p>
<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">images</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">labels</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;labels&quot;</span><span class="p">)</span>
<span class="n">w1</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">var</span><span class="p">(</span><span class="s2">&quot;w1&quot;</span><span class="p">)</span>
<span class="n">b1</span> <span class="o">=</span> <span class="n">pd</span><span class="o">.</span><span class="n">var</span><span class="p">(</span><span class="s2">&quot;b1&quot;</span><span class="p">)</span>
<span class="n">hidden</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">images</span><span class="p">,</span> <span class="n">w</span><span class="o">=</span><span class="n">w1</span><span class="p">,</span> <span class="n">b</span><span class="o">=</span><span class="n">b1</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">hidden</span><span class="p">,</span> <span class="n">labels</span><span class="p">)</span>
</pre></div>
</div>
<p>The above code snippet will create forward operators in <a class="reference external" href="https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/block.md">Block</a>.</p>
</li>
</ol>
<ol>
<li><p class="first">Users create a certain kind of Optimizer with some argument.</p>
<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">optimizer</span> <span class="o">=</span> <span class="n">AdagradOptimizer</span><span class="p">(</span><span class="n">learing_rate</span><span class="o">=</span><span class="mf">0.001</span><span class="p">)</span>
</pre></div>
</div>
</li>
<li><p class="first">Users use the optimizer to <code class="docutils literal"><span class="pre">minimize</span></code> a certain <code class="docutils literal"><span class="pre">cost</span></code> through updating parameters in parameter_list.</p>
<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">opt_op_list</span> <span class="o">=</span> <span class="n">optimizer</span><span class="o">.</span><span class="n">minimize</span><span class="p">(</span><span class="n">cost</span><span class="p">,</span> <span class="n">parameter_list</span><span class="o">=</span><span class="p">[</span><span class="n">w1</span><span class="p">,</span> <span class="n">b1</span><span class="p">])</span>
</pre></div>
</div>
<p>The above code snippet will create gradient and optimization operators in Block. The return value of <code class="docutils literal"><span class="pre">minimize()</span></code> is list of optimization operators that will be run by session.</p>
</li>
<li><p class="first">Users use Session/Executor to run this opt_op_list as target to do training.</p>
<div class="highlight-python"><div class="highlight"><pre><span></span><span class="n">sess</span><span class="o">.</span><span class="n">run</span><span class="p">(</span><span class="n">target</span><span class="o">=</span> <span class="n">opt_op_list</span><span class="p">,</span> <span class="o">...</span><span class="p">)</span>
</pre></div>
</div>
</li>
</ol>
<div class="section" id="optimizer-python-interface">
<span id="optimizer-python-interface"></span><h3>Optimizer Python interface:<a class="headerlink" href="#optimizer-python-interface" title="Permalink to this headline">¶</a></h3>
<div class="highlight-python"><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">Optimizer</span><span class="p">(</span><span class="nb">object</span><span class="p">):</span>
    <span class="sd">&quot;&quot;&quot;Optimizer Base class.</span>

<span class="sd">    &quot;&quot;&quot;</span>

    <span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
        <span class="k">pass</span>

    <span class="k">def</span> <span class="nf">create_optimization_pass</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">parameters_and_grads</span><span class="p">):</span>
        <span class="sd">&quot;&quot;&quot;Add optimization operators to update gradients to variables.</span>

<span class="sd">        Args:</span>
<span class="sd">          parameters_and_grads: a list of (variable, gradient) pair to update.</span>

<span class="sd">        Returns:</span>
<span class="sd">          optmization_op_list: a list of optimization operator that will update parameter using gradient.</span>
<span class="sd">        &quot;&quot;&quot;</span>
        <span class="k">return</span> <span class="bp">None</span>

    <span class="k">def</span> <span class="nf">minimize</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">loss</span><span class="p">,</span> <span class="n">parameter_list</span><span class="p">):</span>
        <span class="sd">&quot;&quot;&quot;Add operations to minimize `loss` by updating `parameter_list`.</span>

<span class="sd">        This method combines interface `append_backward()` and</span>
<span class="sd">        `create_optimization_pass()` into one.</span>
<span class="sd">        &quot;&quot;&quot;</span>
        <span class="n">params_grads</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">create_backward_pass</span><span class="p">(</span><span class="n">loss</span><span class="p">,</span> <span class="n">parameter_list</span><span class="p">)</span>
        <span class="n">update_ops</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">create_optimization_pass</span><span class="p">(</span><span class="n">params_grads</span><span class="p">)</span>
        <span class="k">return</span> <span class="n">update_ops</span>

</pre></div>
</div>
<p>Users can inherit the Optimizer above to create their own Optimizer with some special logic, such as AdagradOptimizer.</p>
</div>
</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="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 type="text/javascript">
      jQuery(function () {
          SphinxRtdTheme.StickyNav.enable();
      });
  </script>
   

</body>
</html>