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
swig_py_paddle_en.html 11.5 KB
Newer Older
Y
Update docs  
Yu Yang 已提交
1 2 3 4 5 6 7 8 
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
  "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">


<html xmlns="http://www.w3.org/1999/xhtml">
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
9 
    <title>Python Prediction API &#8212; PaddlePaddle  documentation</title>
Y
Update docs  
Yu Yang 已提交
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 
    
    <link rel="stylesheet" href="../../_static/classic.css" type="text/css" />
    <link rel="stylesheet" href="../../_static/pygments.css" type="text/css" />
    
    <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://cdn.mathjax.org/mathjax/latest/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>
    <link rel="top" title="PaddlePaddle  documentation" href="../../index.html" />
    <link rel="up" title="User Interface" href="../index.html" />
Y
Upload docs  
Yu Yang 已提交
29 
    <link rel="next" title="Examples and demos" href="../../demo/index.html" />
Y
Update docs  
Yu Yang 已提交
30 31 32 33 34 35 36 37 38 39 40 41 42 
    <link rel="prev" title="Detail Description" href="../cmd_argument/detail_introduction.html" /> 
  </head>
  <body role="document">
    <div class="related" role="navigation" aria-label="related navigation">
      <h3>Navigation</h3>
      <ul>
        <li class="right" style="margin-right: 10px">
          <a href="../../genindex.html" title="General Index"
             accesskey="I">index</a></li>
        <li class="right" >
          <a href="../../py-modindex.html" title="Python Module Index"
             >modules</a> |</li>
        <li class="right" >
Y
Upload docs  
Yu Yang 已提交
43 
          <a href="../../demo/index.html" title="Examples and demos"
Y
Update docs  
Yu Yang 已提交
44 45 46 47 
             accesskey="N">next</a> |</li>
        <li class="right" >
          <a href="../cmd_argument/detail_introduction.html" title="Detail Description"
             accesskey="P">previous</a> |</li>
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
48 49 
        <li class="nav-item nav-item-0"><a href="../../index.html">PaddlePaddle  documentation</a> &#187;</li>
          <li class="nav-item nav-item-1"><a href="../index.html" accesskey="U">User Interface</a> &#187;</li>
Y
Update docs  
Yu Yang 已提交
50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 
      </ul>
    </div>  

    <div class="document">
      <div class="documentwrapper">
        <div class="bodywrapper">
          <div class="body" role="main">
            
  <div class="section" id="python-prediction-api">
<h1>Python Prediction API<a class="headerlink" href="#python-prediction-api" title="Permalink to this headline"></a></h1>
<p>PaddlePaddle offers a set of clean prediction interfaces for python with the help of
SWIG. The main steps of predict values in python are:</p>
<ul class="simple">
<li>Parse training configurations</li>
<li>Construct GradientMachine</li>
<li>Prepare data</li>
<li>Predict</li>
</ul>
<p>Here is a sample python script that shows the typical prediction process for the
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
69 70 71 72 
MNIST classification problem. A complete sample code could be found at
<code class="code docutils literal"><span class="pre">src_root/doc/ui/predict/predict_sample.py</span></code>.</p>
<div class="highlight-python"><div class="highlight"><pre><span></span><span class="kn">from</span> <span class="nn">py_paddle</span> <span class="kn">import</span> <span class="n">swig_paddle</span><span class="p">,</span> <span class="n">DataProviderConverter</span>
<span class="kn">from</span> <span class="nn">paddle.trainer.PyDataProvider2</span> <span class="kn">import</span> <span class="n">dense_vector</span>
Y
Update docs  
Yu Yang 已提交
73 74 75 76 
<span class="kn">from</span> <span class="nn">paddle.trainer.config_parser</span> <span class="kn">import</span> <span class="n">parse_config</span>


<span class="k">def</span> <span class="nf">main</span><span class="p">():</span>
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
77 
    <span class="n">conf</span> <span class="o">=</span> <span class="n">parse_config</span><span class="p">(</span><span class="s2">&quot;./mnist_model/trainer_config.py&quot;</span><span class="p">,</span> <span class="s2">&quot;&quot;</span><span class="p">)</span>
Y
Update docs  
Yu Yang 已提交
78 79 80 81 
    <span class="k">print</span> <span class="n">conf</span><span class="o">.</span><span class="n">data_config</span><span class="o">.</span><span class="n">load_data_args</span>
    <span class="n">network</span> <span class="o">=</span> <span class="n">swig_paddle</span><span class="o">.</span><span class="n">GradientMachine</span><span class="o">.</span><span class="n">createFromConfigProto</span><span class="p">(</span><span class="n">conf</span><span class="o">.</span><span class="n">model_config</span><span class="p">)</span>
    <span class="k">assert</span> <span class="nb">isinstance</span><span class="p">(</span><span class="n">network</span><span class="p">,</span> <span class="n">swig_paddle</span><span class="o">.</span><span class="n">GradientMachine</span><span class="p">)</span>  <span class="c1"># For code hint.</span>
    <span class="n">network</span><span class="o">.</span><span class="n">loadParameters</span><span class="p">(</span><span class="s2">&quot;./mnist_model/&quot;</span><span class="p">)</span>
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
82 
    <span class="n">converter</span> <span class="o">=</span> <span class="n">DataProviderConverter</span><span class="p">([</span><span class="n">dense_vector</span><span class="p">(</span><span class="mi">784</span><span class="p">)])</span>
Y
Update docs  
Yu Yang 已提交
83 84 85 86 87 88 89 90 
    <span class="n">inArg</span> <span class="o">=</span> <span class="n">converter</span><span class="p">(</span><span class="n">TEST_DATA</span><span class="p">)</span>
    <span class="k">print</span> <span class="n">network</span><span class="o">.</span><span class="n">forwardTest</span><span class="p">(</span><span class="n">inArg</span><span class="p">)</span>


<span class="k">if</span> <span class="n">__name__</span> <span class="o">==</span> <span class="s1">&#39;__main__&#39;</span><span class="p">:</span>
    <span class="n">swig_paddle</span><span class="o">.</span><span class="n">initPaddle</span><span class="p">(</span><span class="s2">&quot;--use_gpu=0&quot;</span><span class="p">)</span>
    <span class="n">main</span><span class="p">()</span>
</pre></div>
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
91 
</div>
Y
Update docs  
Yu Yang 已提交
92 93 94 95 
<p>The module that does the most of the job is py_paddle.swig_paddle, it&#8217;s
generated by SWIG and has complete documents, for more details you can use
python&#8217;s <code class="code docutils literal"><span class="pre">help()</span></code> function. Let&#8217;s walk through the above python script:</p>
<ul>
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
96 97 98 
<li><p class="first">At the beginning, use <code class="code docutils literal"><span class="pre">swig_paddle.initPaddle()</span></code> to initialize
PaddlePaddle with command line arguments, for more about command line arguments
see <a class="reference external" href="../cmd_argument/detail_introduction.html">Command Line Arguments</a>.</p>
Y
Update docs  
Yu Yang 已提交
99 
</li>
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
100 101 102 103 
<li><p class="first">Parse the configuration file that is used in training with <code class="code docutils literal"><span class="pre">parse_config()</span></code>.
Because data to predict with always have no label, and output of prediction work
normally is the output layer rather than the cost layer, so you should modify
the configuration file accordingly before using it in the prediction work.</p>
Y
Update docs  
Yu Yang 已提交
104 
</li>
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
105 106 107 108 
<li><p class="first">Create a neural network with
<code class="code docutils literal"><span class="pre">swig_paddle.GradientMachine.createFromConfigproto()</span></code>, which takes the
parsed configuration <code class="code docutils literal"><span class="pre">conf.model_config</span></code> as argument. Then load the
trained parameters from the model with <code class="code docutils literal"><span class="pre">network.loadParameters()</span></code>.</p>
Y
Update docs  
Yu Yang 已提交
109 110 
</li>
<li><dl class="first docutils">
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
111 
<dt>Create a data converter object of utility class <code class="code docutils literal"><span class="pre">DataProviderConverter</span></code>.</dt>
Y
Update docs  
Yu Yang 已提交
112 113 
<dd><ul class="first last simple">
<li>Note: As swig_paddle can only accept C++ matrices, we offer a utility
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
114 115 116 
class DataProviderConverter that can accept the same input data with
PyDataProvider2, for more information please refer to document
of <a class="reference external" href="../data_provider/pydataprovider2.html">PyDataProvider2</a>.</li>
Y
Update docs  
Yu Yang 已提交
117 118 119 120 
</ul>
</dd>
</dl>
</li>
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
121 122 
<li><p class="first">Do the prediction with <code class="code docutils literal"><span class="pre">forwardTest()</span></code>, which takes the converted
input data and outputs the activations of the output layer.</p>
Y
Update docs  
Yu Yang 已提交
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 
</li>
</ul>
<p>Here is a typical output:</p>
<div class="highlight-text"><div class="highlight"><pre><span></span>[{&#39;id&#39;: None, &#39;value&#39;: array([[  5.53018653e-09,   1.12194102e-05,   1.96644767e-09,
      1.43630644e-02,   1.51111044e-13,   9.85625684e-01,
      2.08823112e-10,   2.32777140e-08,   2.00186201e-09,
      1.15501715e-08],
   [  9.99982715e-01,   1.27787406e-10,   1.72296313e-05,
      1.49316648e-09,   1.36540484e-11,   6.93137714e-10,
      2.70634608e-08,   3.48565123e-08,   5.25639710e-09,
      4.48684503e-08]], dtype=float32)}]
</pre></div>
</div>
<p><code class="code docutils literal"><span class="pre">value</span></code> is the output of the output layer, each row represents result of
the corresponding row in the input data, each element represents activation of
the corresponding neuron in the output layer.</p>
</div>


          </div>
        </div>
      </div>
      <div class="sphinxsidebar" role="navigation" aria-label="main navigation">
        <div class="sphinxsidebarwrapper">
  <h4>Previous topic</h4>
  <p class="topless"><a href="../cmd_argument/detail_introduction.html"
                        title="previous chapter">Detail Description</a></p>
  <h4>Next topic</h4>
Y
Upload docs  
Yu Yang 已提交
151 152 
  <p class="topless"><a href="../../demo/index.html"
                        title="next chapter">Examples and demos</a></p>
Y
Update docs  
Yu Yang 已提交
153 154 155 156 157 158 159 160 161 162 
  <div role="note" aria-label="source link">
    <h3>This Page</h3>
    <ul class="this-page-menu">
      <li><a href="../../_sources/ui/predict/swig_py_paddle_en.txt"
            rel="nofollow">Show Source</a></li>
    </ul>
   </div>
<div id="searchbox" style="display: none" role="search">
  <h3>Quick search</h3>
    <form class="search" action="../../search.html" method="get">
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
163 164 
      <div><input type="text" name="q" /></div>
      <div><input type="submit" value="Go" /></div>
Y
Update docs  
Yu Yang 已提交
165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 
      <input type="hidden" name="check_keywords" value="yes" />
      <input type="hidden" name="area" value="default" />
    </form>
</div>
<script type="text/javascript">$('#searchbox').show(0);</script>
        </div>
      </div>
      <div class="clearer"></div>
    </div>
    <div class="related" role="navigation" aria-label="related navigation">
      <h3>Navigation</h3>
      <ul>
        <li class="right" style="margin-right: 10px">
          <a href="../../genindex.html" title="General Index"
             >index</a></li>
        <li class="right" >
          <a href="../../py-modindex.html" title="Python Module Index"
             >modules</a> |</li>
        <li class="right" >
Y
Upload docs  
Yu Yang 已提交
184 
          <a href="../../demo/index.html" title="Examples and demos"
Y
Update docs  
Yu Yang 已提交
185 186 187 188 
             >next</a> |</li>
        <li class="right" >
          <a href="../cmd_argument/detail_introduction.html" title="Detail Description"
             >previous</a> |</li>
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
189 190 
        <li class="nav-item nav-item-0"><a href="../../index.html">PaddlePaddle  documentation</a> &#187;</li>
          <li class="nav-item nav-item-1"><a href="../index.html" >User Interface</a> &#187;</li>
Y
Update docs  
Yu Yang 已提交
191 192 193 
      </ul>
    </div>
    <div class="footer" role="contentinfo">
T
Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5  
Travis CI 已提交
194 195 
        &#169; Copyright 2016, PaddlePaddle developers.
      Created using <a href="http://sphinx-doc.org/">Sphinx</a> 1.4.6.
Y
Update docs  
Yu Yang 已提交
196 197 198 
    </div>
  </body>
</html>