@@ -496,6 +497,51 @@ index that reader returns.</li>
<dd><p>Event On One Batch Training Complete.</p>
</dd></dl>
</div>
<divclass="section"id="inference">
<h2>Inference<aclass="headerlink"href="#inference"title="Permalink to this headline">¶</a></h2>
<dlclass="function">
<dtid="paddle.v2.infer">
<codeclass="descclassname">paddle.v2.</code><codeclass="descname">infer</code><spanclass="sig-paren">(</span><em>output</em>, <em>parameters</em>, <em>input=None</em>, <em>batch_size=None</em>, <em>reader=None</em>, <em>feeding=None</em>, <em>field='value'</em><spanclass="sig-paren">)</span><aclass="headerlink"href="#paddle.v2.infer"title="Permalink to this definition">¶</a></dt>
<dd><p>Infer a neural network by given neural network output and parameters. The
user should pass either a batch of input data or reader method.</p>
<li><strong>output</strong> (<em>paddle.v2.config_base.Layer</em>) – output of the neural network that would be inferred</li>
<li><strong>parameters</strong> (<aclass="reference internal"href="#paddle.v2.parameters.Parameters"title="paddle.v2.parameters.Parameters"><em>paddle.v2.parameters.Parameters</em></a>) – parameters of the neural network.</li>
<li><strong>input</strong> (<em>collections.Iterable</em>) – input data batch. Should be a python iterable object, and each
element is the data batch.</li>
<li><strong>batch_size</strong> (<em>int</em>) – the batch size when perform inference. Default is the
length of input.</li>
<li><strong>reader</strong> (<em>callable</em>) – input data reader creator in batch. If this field is set, the
<cite>input</cite> and <cite>batch_size</cite> will be ignored.</li>
<li><strong>feeding</strong>– Reader dictionary. Default could generate from input
value.</li>
<li><strong>field</strong> (<em>str</em>) – The prediction field. It should in [<cite>value</cite>, <cite>ids</cite>]. <cite>value</cite>
means return the prediction probabilities, <cite>ids</cite> means return
the prediction labels. Default is <cite>value</cite></li>
<li><strong>output</strong> (<em>paddle.v2.config_base.Layer</em>) – output of the neural network that would be inferred</li>
<li><strong>parameters</strong> (<aclass="reference internal"href="#paddle.v2.parameters.Parameters"title="paddle.v2.parameters.Parameters"><em>paddle.v2.parameters.Parameters</em></a>) – parameters of the neural network.</li>
<li><strong>input</strong> (<em>collections.Iterable</em>) – input data batch. Should be a python iterable object, and each
element is the data batch.</li>
<li><strong>batch_size</strong> (<em>int</em>) – the batch size when perform inference. Default is the
length of input.</li>
<li><strong>reader</strong> (<em>callable</em>) – input data reader creator in batch. If this field is set, the
<cite>input</cite> and <cite>batch_size</cite> will be ignored.</li>
<li><strong>feeding</strong>– Reader dictionary. Default could generate from input
value.</li>
<li><strong>field</strong> (<em>str</em>) – The prediction field. It should in [<cite>value</cite>, <cite>ids</cite>]. <cite>value</cite>
means return the prediction probabilities, <cite>ids</cite> means return
the prediction labels. Default is <cite>value</cite></li>