Deploy to GitHub Pages: 2f82d72e

doc/ui/api/trainer_config_helpers/networks.html

@@ -751,14 +751,13 @@ compute attention weight.</li>
 <dl class="function">
 <dt>
 <code class="descclassname">paddle.trainer_config_helpers.networks.</code><code class="descname">outputs</code><span class="sig-paren">(</span><em>layers</em>, <em>*args</em><span class="sig-paren">)</span></dt>
-<dd><p>Declare the end of network. Currently it will only calculate the
-input/output order of network. It will calculate the predict network or
-train network&#8217;s output automatically.</p>
+<dd><p>Declare the outputs of network. If user have not defined the inputs of
+network, this method will calculate the input order by dfs travel.</p>
 <table class="docutils field-list" frame="void" rules="none">
 <col class="field-name" />
 <col class="field-body" />
 <tbody valign="top">
-<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>layers</strong> (<em>list|tuple|LayerOutput</em>) &#8211; </td>
+<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>layers</strong> (<em>list|tuple|LayerOutput</em>) &#8211; Output layers.</td>
 </tr>
 <tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"></td>
 </tr>

doc/ui/data_provider/pydataprovider2.html

@@ -199,10 +199,10 @@ the <code class="code docutils literal"><span class="pre">dataprovider</span></c
 
 
 <span class="c1"># Define a py data provider</span>
-<span class="nd">@provider</span><span class="p">(</span><span class="n">input_types</span><span class="o">=</span><span class="p">[</span>
-    <span class="n">dense_vector</span><span class="p">(</span><span class="mi">28</span> <span class="o">*</span> <span class="mi">28</span><span class="p">),</span>
-    <span class="n">integer_value</span><span class="p">(</span><span class="mi">10</span><span class="p">)</span>
-<span class="p">])</span>
+<span class="nd">@provider</span><span class="p">(</span><span class="n">input_types</span><span class="o">=</span><span class="p">{</span>
+    <span class="s1">&#39;pixel&#39;</span><span class="p">:</span> <span class="n">dense_vector</span><span class="p">(</span><span class="mi">28</span> <span class="o">*</span> <span class="mi">28</span><span class="p">),</span>
+    <span class="s1">&#39;label&#39;</span><span class="p">:</span> <span class="n">integer_value</span><span class="p">(</span><span class="mi">10</span><span class="p">)</span>
+<span class="p">})</span>
 <span class="k">def</span> <span class="nf">process</span><span class="p">(</span><span class="n">settings</span><span class="p">,</span> <span class="n">filename</span><span class="p">):</span>  <span class="c1"># settings is not used currently.</span>
     <span class="n">f</span> <span class="o">=</span> <span class="nb">open</span><span class="p">(</span><span class="n">filename</span><span class="p">,</span> <span class="s1">&#39;r&#39;</span><span class="p">)</span>  <span class="c1"># open one of training file</span>
 
@@ -217,7 +217,7 @@ the <code class="code docutils literal"><span class="pre">dataprovider</span></c
             <span class="n">pixels_float</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="nb">float</span><span class="p">(</span><span class="n">each_pixel_str</span><span class="p">))</span>
 
         <span class="c1"># give data to paddle.</span>
-        <span class="k">yield</span> <span class="p">{</span> <span class="s2">&quot;pixel&quot;</span><span class="p">:</span> <span class="n">pixels_float</span><span class="p">,</span> <span class="s1">&#39;label&#39;</span><span class="p">:</span> <span class="nb">int</span><span class="p">(</span><span class="n">label</span><span class="p">)</span> <span class="p">}</span>
+        <span class="k">yield</span> <span class="p">{</span><span class="s2">&quot;pixel&quot;</span><span class="p">:</span> <span class="n">pixels_float</span><span class="p">,</span> <span class="s1">&#39;label&#39;</span><span class="p">:</span> <span class="nb">int</span><span class="p">(</span><span class="n">label</span><span class="p">)}</span>
 
     <span class="n">f</span><span class="o">.</span><span class="n">close</span><span class="p">()</span>  <span class="c1"># close file</span>
 </pre></div>
@@ -355,7 +355,7 @@ Please refer to the following section reference for details.</p>
 <h3>&#64;provider<a class="headerlink" href="#provider" title="Permalink to this headline">¶</a></h3>
 <dl class="function">
 <dt id="paddle.trainer.PyDataProvider2.provider">
-<code class="descclassname">paddle.trainer.PyDataProvider2.</code><code class="descname">provider</code><span class="sig-paren">(</span><em>input_types=None</em>, <em>should_shuffle=None</em>, <em>pool_size=-1</em>, <em>min_pool_size=-1</em>, <em>can_over_batch_size=True</em>, <em>calc_batch_size=None</em>, <em>cache=0</em>, <em>check=False</em>, <em>check_fail_continue=False</em>, <em>use_dynamic_order=True</em>, <em>init_hook=None</em>, <em>**kwargs</em><span class="sig-paren">)</span><a class="headerlink" href="#paddle.trainer.PyDataProvider2.provider" title="Permalink to this definition">¶</a></dt>
+<code class="descclassname">paddle.trainer.PyDataProvider2.</code><code class="descname">provider</code><span class="sig-paren">(</span><em>input_types=None</em>, <em>should_shuffle=None</em>, <em>pool_size=-1</em>, <em>min_pool_size=-1</em>, <em>can_over_batch_size=True</em>, <em>calc_batch_size=None</em>, <em>cache=0</em>, <em>check=False</em>, <em>check_fail_continue=False</em>, <em>init_hook=None</em>, <em>**kwargs</em><span class="sig-paren">)</span><a class="headerlink" href="#paddle.trainer.PyDataProvider2.provider" title="Permalink to this definition">¶</a></dt>
 <dd><p>Provider decorator. Use it to make a function into PyDataProvider2 object.
 In this function, user only need to get each sample for some train/test
 file.</p>
@@ -373,8 +373,13 @@ file.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
-<li><strong>input_types</strong> (<em>list|tuple</em>) &#8211; Specify the input types, can also be set in init_hook.
-It is a list of InputType object. For example, input_types=                         [dense_vector(9), integer_value(2)].</li>
+<li><strong>input_types</strong> (<em>list|tuple|dict</em>) &#8211; Specify the input types, can also be set in init_hook.
+It could be a list of InputType object. For example,
+input_types=[dense_vector(9), integer_value(2)]. Or user
+can set a dict of InputType object, which key is
+data_layer&#8217;s name. For example, input_types=                        {&#8216;img&#8217;: img_features, &#8216;label&#8217;: label}. when using dict of
+InputType, user could yield a dict of feature values, which
+key is also data_layer&#8217;s name.</li>
 <li><strong>should_shuffle</strong> (<em>bool</em>) &#8211; True if data should shuffle. Pass None means shuffle
 when is training and not to shuffle when is testing.</li>
 <li><strong>pool_size</strong> (<em>int</em>) &#8211; Max number of sample in data pool.</li>
@@ -409,10 +414,6 @@ for debug. Default is disabled.</li>
 <li><strong>check_fail_continue</strong> (<em>bool</em>) &#8211; Continue train or not when check failed. Just
 drop the wrong format data when it is True. Has
 no effect when check set to False.</li>
-<li><strong>use_dynamic_order</strong> (<em>bool</em>) &#8211; Allow provider to yield a dictionary object, whose
-key is a input data layer name, and value is the
-feature value. The tuples are still allowed when
-use_dynmaic_order is True.</li>
 </ul>
 </td>
 </tr>

doc_cn/_sources/ui/data_provider/pydataprovider2.txt

@@ -141,8 +141,6 @@ DataProvider创建的时候执行。这个初始化函数具有如下参数:
    是一个batch size，但是有时为了计算均衡性，可以将一条数据设置成多个batch size
 *  cache 是数据缓存的策略，参考 `cache`_
 *  init_hook 是初始化时调用的函数，参考 `init_hook`_
-*  use_dynamic_order 如果是true的话，可以返回一个dict，key是data_layer的名字，value是特征值。同时，也可以
-   返回一个list或者tuple。如果是false的话，只能够返回list或者tuple
 *  check 设置成true的话，会根据input_types检查数据的合法性。
 *  check_fail_continue 如果设置成true的话，即使在check中数据不合法，也会扔到这条数据，继续训练。 如果
    check是false的话，没有作用。

doc_cn/ui/data_provider/pydataprovider2.html

@@ -189,10 +189,10 @@ process函数调用多次 <code class="code docutils literal"><span class="pre">
 
 
 <span class="c1"># Define a py data provider</span>
-<span class="nd">@provider</span><span class="p">(</span><span class="n">input_types</span><span class="o">=</span><span class="p">[</span>
-    <span class="n">dense_vector</span><span class="p">(</span><span class="mi">28</span> <span class="o">*</span> <span class="mi">28</span><span class="p">),</span>
-    <span class="n">integer_value</span><span class="p">(</span><span class="mi">10</span><span class="p">)</span>
-<span class="p">])</span>
+<span class="nd">@provider</span><span class="p">(</span><span class="n">input_types</span><span class="o">=</span><span class="p">{</span>
+    <span class="s1">&#39;pixel&#39;</span><span class="p">:</span> <span class="n">dense_vector</span><span class="p">(</span><span class="mi">28</span> <span class="o">*</span> <span class="mi">28</span><span class="p">),</span>
+    <span class="s1">&#39;label&#39;</span><span class="p">:</span> <span class="n">integer_value</span><span class="p">(</span><span class="mi">10</span><span class="p">)</span>
+<span class="p">})</span>
 <span class="k">def</span> <span class="nf">process</span><span class="p">(</span><span class="n">settings</span><span class="p">,</span> <span class="n">filename</span><span class="p">):</span>  <span class="c1"># settings is not used currently.</span>
     <span class="n">f</span> <span class="o">=</span> <span class="nb">open</span><span class="p">(</span><span class="n">filename</span><span class="p">,</span> <span class="s1">&#39;r&#39;</span><span class="p">)</span>  <span class="c1"># open one of training file</span>
 
@@ -207,7 +207,7 @@ process函数调用多次 <code class="code docutils literal"><span class="pre">
             <span class="n">pixels_float</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="nb">float</span><span class="p">(</span><span class="n">each_pixel_str</span><span class="p">))</span>
 
         <span class="c1"># give data to paddle.</span>
-        <span class="k">yield</span> <span class="p">{</span> <span class="s2">&quot;pixel&quot;</span><span class="p">:</span> <span class="n">pixels_float</span><span class="p">,</span> <span class="s1">&#39;label&#39;</span><span class="p">:</span> <span class="nb">int</span><span class="p">(</span><span class="n">label</span><span class="p">)</span> <span class="p">}</span>
+        <span class="k">yield</span> <span class="p">{</span><span class="s2">&quot;pixel&quot;</span><span class="p">:</span> <span class="n">pixels_float</span><span class="p">,</span> <span class="s1">&#39;label&#39;</span><span class="p">:</span> <span class="nb">int</span><span class="p">(</span><span class="n">label</span><span class="p">)}</span>
 
     <span class="n">f</span><span class="o">.</span><span class="n">close</span><span class="p">()</span>  <span class="c1"># close file</span>
 </pre></div>
@@ -340,8 +340,6 @@ DataProvider创建的时候执行。这个初始化函数具有如下参数:</p>
 是一个batch size，但是有时为了计算均衡性，可以将一条数据设置成多个batch size</li>
 <li>cache 是数据缓存的策略，参考 <a class="reference internal" href="#cache">cache</a></li>
 <li>init_hook 是初始化时调用的函数，参考 <a class="reference internal" href="#init-hook">init_hook</a></li>
-<li>use_dynamic_order 如果是true的话，可以返回一个dict，key是data_layer的名字，value是特征值。同时，也可以
-返回一个list或者tuple。如果是false的话，只能够返回list或者tuple</li>
 <li>check 设置成true的话，会根据input_types检查数据的合法性。</li>
 <li>check_fail_continue 如果设置成true的话，即使在check中数据不合法，也会扔到这条数据，继续训练。 如果
 check是false的话，没有作用。</li>