cpp_data_feeding.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>C++ Data Feeding &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 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">
          
            
                <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="../build_and_install/index_en.html">Install and Build</a></li>
<li class="toctree-l1"><a class="reference internal" href="../howto/index_en.html">HOW TO</a></li>
<li class="toctree-l1"><a class="reference internal" href="../dev/index_en.html">Development</a></li>
<li class="toctree-l1"><a class="reference internal" href="../faq/index_en.html">FAQ</a></li>
</ul>

            
        </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>C++ Data Feeding</li>
      <li class="wy-breadcrumbs-aside">
        
          
            <a href="../_sources/design/cpp_data_feeding.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="c-data-feeding">
<span id="c-data-feeding"></span><h1>C++ Data Feeding<a class="headerlink" href="#c-data-feeding" title="Permalink to this headline">¶</a></h1>
<p>In training with Paddle V2 API, data feeding wholly dependents on Python code. To get rid of the Python environment and achieve the goal of &#8220;wrapping the whole training by a while loop op&#8221; in Paddle Fluid, a C++ data feeding mechanism is required.</p>
<p>In this document we show the fundamental design of C++ data feeding process, which includes the data reading, shuffling and batching.</p>
<div class="section" id="reader">
<span id="reader"></span><h2>Reader<a class="headerlink" href="#reader" title="Permalink to this headline">¶</a></h2>
<p>A new concept named &#8216;Reader&#8217; is introduced. <code class="docutils literal"><span class="pre">Reader</span></code> is a series of inherited classes which can be hold by our <code class="docutils literal"><span class="pre">Variable</span></code> and they are used to read or process file data.</p>
<div class="section" id="readerbase">
<span id="readerbase"></span><h3><code class="docutils literal"><span class="pre">ReaderBase</span></code><a class="headerlink" href="#readerbase" title="Permalink to this headline">¶</a></h3>
<p><code class="docutils literal"><span class="pre">ReaderBase</span></code> is the abstract base class of all readers. It defines the all readers&#8217; interfaces.</p>
<div class="highlight-cpp"><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">ReaderBase</span> <span class="p">{</span>
 <span class="k">public</span><span class="o">:</span>
  <span class="k">explicit</span> <span class="n">ReaderBase</span><span class="p">(</span><span class="k">const</span> <span class="n">std</span><span class="o">::</span><span class="n">vector</span><span class="o">&lt;</span><span class="n">DDim</span><span class="o">&gt;&amp;</span> <span class="n">shapes</span><span class="p">)</span> <span class="o">:</span> <span class="n">shapes_</span><span class="p">(</span><span class="n">shapes</span><span class="p">)</span> <span class="p">{</span>
    <span class="n">PADDLE_ENFORCE</span><span class="p">(</span><span class="o">!</span><span class="n">shapes_</span><span class="p">.</span><span class="n">empty</span><span class="p">());</span>
  <span class="p">}</span>
  <span class="c1">// Read the next batch of data. (A &#39;batch&#39; can be only one instance)</span>
  <span class="k">virtual</span> <span class="kt">void</span> <span class="n">ReadNext</span><span class="p">(</span><span class="n">std</span><span class="o">::</span><span class="n">vector</span><span class="o">&lt;</span><span class="n">LoDTensor</span><span class="o">&gt;*</span> <span class="n">out</span><span class="p">)</span> <span class="o">=</span> <span class="mi">0</span><span class="p">;</span>
  <span class="c1">// Show whether the next bacth exists.</span>
  <span class="k">virtual</span> <span class="kt">bool</span> <span class="nf">HasNext</span><span class="p">()</span> <span class="k">const</span> <span class="o">=</span> <span class="mi">0</span><span class="p">;</span>
  
  <span class="c1">// Reinitialize the reader and read the file from the begin.</span>
  <span class="k">virtual</span> <span class="kt">void</span> <span class="nf">ReInit</span><span class="p">()</span> <span class="o">=</span> <span class="mi">0</span><span class="p">;</span>
  
  <span class="c1">// Get a certain read in data&#39;s shape.</span>
  <span class="n">DDim</span> <span class="nf">shape</span><span class="p">(</span><span class="kt">size_t</span> <span class="n">idx</span><span class="p">)</span> <span class="k">const</span><span class="p">;</span>
  <span class="c1">// Get shapes of all read in data.</span>
  <span class="n">std</span><span class="o">::</span><span class="n">vector</span><span class="o">&lt;</span><span class="n">DDim</span><span class="o">&gt;</span> <span class="n">shapes</span><span class="p">()</span> <span class="k">const</span> <span class="p">{</span> <span class="k">return</span> <span class="n">shapes_</span><span class="p">;</span> <span class="p">}</span>
  <span class="c1">// Set shapes of read in data.</span>
  <span class="kt">void</span> <span class="n">set_shapes</span><span class="p">(</span><span class="k">const</span> <span class="n">std</span><span class="o">::</span><span class="n">vector</span><span class="o">&lt;</span><span class="n">DDim</span><span class="o">&gt;&amp;</span> <span class="n">shapes</span><span class="p">)</span> <span class="p">{</span> <span class="n">shapes_</span> <span class="o">=</span> <span class="n">shapes</span><span class="p">;</span> <span class="p">}</span>

  <span class="k">virtual</span> <span class="o">~</span><span class="n">ReaderBase</span><span class="p">()</span> <span class="p">{}</span>

 <span class="k">protected</span><span class="o">:</span>
  <span class="n">std</span><span class="o">::</span><span class="n">vector</span><span class="o">&lt;</span><span class="n">DDim</span><span class="o">&gt;</span> <span class="n">shapes_</span><span class="p">;</span>
<span class="p">};</span>
</pre></div>
</div>
</div>
<div class="section" id="filereader-and-decoratedreader">
<span id="filereader-and-decoratedreader"></span><h3><code class="docutils literal"><span class="pre">FileReader</span></code> and <code class="docutils literal"><span class="pre">DecoratedReader</span></code><a class="headerlink" href="#filereader-and-decoratedreader" title="Permalink to this headline">¶</a></h3>
<p>These two classes are derived from the <code class="docutils literal"><span class="pre">ReaderBase</span></code> and will further be derived by respective specific readers. That is to say, in our design, there are two kinds of readers: file readers and decorated readers. A file reader reads from a file of some specific format, and yield only one instance of data at a time. e.g. RecordIO reader, jpg reader, .... A decorated reader takes another reader(both file reader and decorated reader are OK) as its &#8216;underlying reader&#8217;. It gets data from its underlying reader, does some process on them(shuffling, or batching), then yields processed data. The output data of a decorated reader can be a single instance or a batch. <code class="docutils literal"><span class="pre">ShuffleReader</span></code> and <code class="docutils literal"><span class="pre">BatchReader</span></code> are both decorated readers.</p>
<p>All the readers share exactly the same interfaces defined in <code class="docutils literal"><span class="pre">ReaderBase</span></code>. So they can be decorated for more than one time: We can <strong>shuffle</strong> a reader&#8217;s outputs and then <strong>batch</strong> the shuffle outputs. The interface consistency also allows related ops use readers without knowing what they are exactly.</p>
</div>
<div class="section" id="readerholder">
<span id="readerholder"></span><h3><code class="docutils literal"><span class="pre">ReaderHolder</span></code><a class="headerlink" href="#readerholder" title="Permalink to this headline">¶</a></h3>
<p>Different readers belong to different class types. It leads to a problem: How can we drop them into <code class="docutils literal"><span class="pre">Variable</span></code>s and fetch them out by a unified method? For example, if a Variable holds a <code class="docutils literal"><span class="pre">BatchReader</span></code>, we can not get it by the following code:</p>
<div class="highlight-cpp"><div class="highlight"><pre><span></span><span class="n">var</span><span class="o">-&gt;</span><span class="n">Get</span><span class="o">&lt;</span><span class="n">ReaderBase</span><span class="o">&gt;</span><span class="p">(</span><span class="s">&quot;batch_reader&quot;</span><span class="p">);</span>
</pre></div>
</div>
<p>we have to write:</p>
<div class="highlight-cpp"><div class="highlight"><pre><span></span><span class="n">var</span><span class="o">-&gt;</span><span class="n">Get</span><span class="o">&lt;</span><span class="n">BatchReader</span><span class="o">&gt;</span><span class="p">(</span><span class="s">&quot;batch_reader&quot;</span><span class="p">);</span>
</pre></div>
</div>
<p>This requires each time getting a reader from a variable we must know the reader&#8217;s type exactly. It is nearly impossible.</p>
<p>To solve this problem, we introduce <code class="docutils literal"><span class="pre">ReaderHolder</span></code> as a wrapper. It acts as an empty decorator of <code class="docutils literal"><span class="pre">ReaderBase</span></code>, which erases reader&#8217;s type. With <code class="docutils literal"><span class="pre">ReaderHolder</span></code> we are able to fetch all types of readers by <code class="docutils literal"><span class="pre">var-&gt;Get&lt;ReaderHolder&gt;(&quot;...&quot;)</span></code> and regard the obtained object as a reader.</p>
</div>
</div>
<div class="section" id="related-operators">
<span id="related-operators"></span><h2>Related Operators<a class="headerlink" href="#related-operators" title="Permalink to this headline">¶</a></h2>
<p>To create and invoke readers, some now ops are introduced:</p>
<div class="section" id="createreaderop">
<span id="createreaderop"></span><h3><code class="docutils literal"><span class="pre">CreateReaderOp</span></code><a class="headerlink" href="#createreaderop" title="Permalink to this headline">¶</a></h3>
<p>Each reader has its creating op. File readers&#8217; creating ops have no input and yield the created file reader as its output. Decorated readers&#8217; creating ops take the underlying readers as inputs and then yield new decorated readers.</p>
</div>
<div class="section" id="readop">
<span id="readop"></span><h3><code class="docutils literal"><span class="pre">ReadOp</span></code><a class="headerlink" href="#readop" title="Permalink to this headline">¶</a></h3>
<p>A reader is only a Variable. It cannot trigger the reading process by itself. So we add the <code class="docutils literal"><span class="pre">ReadOp</span></code> to execute it. A <code class="docutils literal"><span class="pre">ReadOp</span></code> takes a reader Variable as its input. Each time it runs, it invokes the reader‘s <code class="docutils literal"><span class="pre">ReadNext()</span></code> function and gets a new batch of data(or only one instance of data, if we use file reader directly). The output data of a reader are in the form of <code class="docutils literal"><span class="pre">std::vector&lt;LoDTenosr&gt;</span></code>, so the <code class="docutils literal"><span class="pre">ReadOp</span></code> also needs to split the vector and move LoDTensors to their respective output Variables.</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>