Deploy to GitHub Pages: d8b923ab

develop/doc/api/v2/fluid/layers.html

@@ -309,10 +309,11 @@ layer.</li>
 <h2>embedding<a class="headerlink" href="#embedding" title="Permalink to this headline">¶</a></h2>
 <dl class="function">
 <dt>
-<code class="descclassname">paddle.v2.fluid.layers.</code><code class="descname">embedding</code><span class="sig-paren">(</span><em>input</em>, <em>size</em>, <em>is_sparse=False</em>, <em>param_attr=None</em>, <em>dtype='float32'</em><span class="sig-paren">)</span></dt>
+<code class="descclassname">paddle.v2.fluid.layers.</code><code class="descname">embedding</code><span class="sig-paren">(</span><em>input</em>, <em>size</em>, <em>is_sparse=False</em>, <em>padding_idx=None</em>, <em>param_attr=None</em>, <em>dtype='float32'</em><span class="sig-paren">)</span></dt>
 <dd><p><strong>Embedding Layer</strong></p>
-<p>This layer is used to lookup a vector of IDs, provided by <em>input</em>, in a lookup table.
-The result of this lookup is the embedding of each ID in the <em>input</em>.</p>
+<p>This layer is used to lookup embeddings of IDs, provided by <code class="xref py py-attr docutils literal"><span class="pre">input</span></code>, in
+a lookup table. The result of this lookup is the embedding of each ID in the
+<code class="xref py py-attr docutils literal"><span class="pre">input</span></code>.</p>
 <p>All the input variables are passed in as local variables to the LayerHelper
 constructor.</p>
 <table class="docutils field-list" frame="void" rules="none">
@@ -320,9 +321,16 @@ constructor.</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 simple">
-<li><strong>input</strong> (<em>Variable</em>) &#8211; Input to the function</li>
-<li><strong>size</strong> (<em>tuple|list|None</em>) &#8211; Shape of the look up table parameter</li>
-<li><strong>is_sparse</strong> (<em>bool</em>) &#8211; Boolean flag that specifying whether the input is sparse</li>
+<li><strong>input</strong> (<em>Variable</em>) &#8211; The tensor variable containing the IDs.</li>
+<li><strong>size</strong> (<em>tuple|list</em>) &#8211; The shape of the look up table parameter. It should
+have two elements which indicate the size of the dictionary of
+embeddings and the size of each embedding vector respectively.</li>
+<li><strong>is_sparse</strong> (<em>bool</em>) &#8211; The flag indicating whether to use sparse update.</li>
+<li><strong>padding_idx</strong> (<em>int|long|None</em>) &#8211; If <code class="xref py py-attr docutils literal"><span class="pre">None</span></code>, it makes no effect to lookup.
+Otherwise the given <code class="xref py py-attr docutils literal"><span class="pre">padding_idx</span></code> indicates padding the output
+with zeros whenever lookup encounters it in <code class="xref py py-attr docutils literal"><span class="pre">input</span></code>. If
+<span class="math">\(padding_idx &lt; 0\)</span>, the padding_idx to use in lookup is
+<span class="math">\(size[0] + dim\)</span>.</li>
 <li><strong>param_attr</strong> (<em>ParamAttr</em>) &#8211; Parameters for this layer</li>
 <li><strong>dtype</strong> (<em>np.dtype|core.DataType|str</em>) &#8211; The type of data : float32, float_16, int etc</li>
 </ul>
@@ -1044,10 +1052,11 @@ that need to be summed up.</td>
 <h2>assign<a class="headerlink" href="#assign" title="Permalink to this headline">¶</a></h2>
 <dl class="function">
 <dt>
-<code class="descclassname">paddle.v2.fluid.layers.</code><code class="descname">embedding</code><span class="sig-paren">(</span><em>input</em>, <em>size</em>, <em>is_sparse=False</em>, <em>param_attr=None</em>, <em>dtype='float32'</em><span class="sig-paren">)</span></dt>
+<code class="descclassname">paddle.v2.fluid.layers.</code><code class="descname">embedding</code><span class="sig-paren">(</span><em>input</em>, <em>size</em>, <em>is_sparse=False</em>, <em>padding_idx=None</em>, <em>param_attr=None</em>, <em>dtype='float32'</em><span class="sig-paren">)</span></dt>
 <dd><p><strong>Embedding Layer</strong></p>
-<p>This layer is used to lookup a vector of IDs, provided by <em>input</em>, in a lookup table.
-The result of this lookup is the embedding of each ID in the <em>input</em>.</p>
+<p>This layer is used to lookup embeddings of IDs, provided by <code class="xref py py-attr docutils literal"><span class="pre">input</span></code>, in
+a lookup table. The result of this lookup is the embedding of each ID in the
+<code class="xref py py-attr docutils literal"><span class="pre">input</span></code>.</p>
 <p>All the input variables are passed in as local variables to the LayerHelper
 constructor.</p>
 <table class="docutils field-list" frame="void" rules="none">
@@ -1055,9 +1064,16 @@ constructor.</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 simple">
-<li><strong>input</strong> (<em>Variable</em>) &#8211; Input to the function</li>
-<li><strong>size</strong> (<em>tuple|list|None</em>) &#8211; Shape of the look up table parameter</li>
-<li><strong>is_sparse</strong> (<em>bool</em>) &#8211; Boolean flag that specifying whether the input is sparse</li>
+<li><strong>input</strong> (<em>Variable</em>) &#8211; The tensor variable containing the IDs.</li>
+<li><strong>size</strong> (<em>tuple|list</em>) &#8211; The shape of the look up table parameter. It should
+have two elements which indicate the size of the dictionary of
+embeddings and the size of each embedding vector respectively.</li>
+<li><strong>is_sparse</strong> (<em>bool</em>) &#8211; The flag indicating whether to use sparse update.</li>
+<li><strong>padding_idx</strong> (<em>int|long|None</em>) &#8211; If <code class="xref py py-attr docutils literal"><span class="pre">None</span></code>, it makes no effect to lookup.
+Otherwise the given <code class="xref py py-attr docutils literal"><span class="pre">padding_idx</span></code> indicates padding the output
+with zeros whenever lookup encounters it in <code class="xref py py-attr docutils literal"><span class="pre">input</span></code>. If
+<span class="math">\(padding_idx &lt; 0\)</span>, the padding_idx to use in lookup is
+<span class="math">\(size[0] + dim\)</span>.</li>
 <li><strong>param_attr</strong> (<em>ParamAttr</em>) &#8211; Parameters for this layer</li>
 <li><strong>dtype</strong> (<em>np.dtype|core.DataType|str</em>) &#8211; The type of data : float32, float_16, int etc</li>
 </ul>

develop/doc/api/v2/fluid/nets.html

@@ -230,7 +230,7 @@
 <h2>img_conv_group<a class="headerlink" href="#img-conv-group" title="Permalink to this headline">¶</a></h2>
 <dl class="function">
 <dt>
-<code class="descclassname">paddle.v2.fluid.nets.</code><code class="descname">img_conv_group</code><span class="sig-paren">(</span><em>input</em>, <em>conv_num_filter</em>, <em>pool_size</em>, <em>conv_padding=1</em>, <em>conv_filter_size=3</em>, <em>conv_act=None</em>, <em>param_attr=None</em>, <em>conv_with_batchnorm=False</em>, <em>conv_batchnorm_drop_rate=None</em>, <em>pool_stride=1</em>, <em>pool_type=None</em>, <em>use_cudnn=True</em><span class="sig-paren">)</span></dt>
+<code class="descclassname">paddle.v2.fluid.nets.</code><code class="descname">img_conv_group</code><span class="sig-paren">(</span><em>input</em>, <em>conv_num_filter</em>, <em>pool_size</em>, <em>conv_padding=1</em>, <em>conv_filter_size=3</em>, <em>conv_act=None</em>, <em>param_attr=None</em>, <em>conv_with_batchnorm=False</em>, <em>conv_batchnorm_drop_rate=0.0</em>, <em>pool_stride=1</em>, <em>pool_type=None</em>, <em>use_cudnn=True</em><span class="sig-paren">)</span></dt>
 <dd><p>Image Convolution Group, Used for vgg net.</p>
 </dd></dl>
 

develop/doc/operators.json

@@ -3440,6 +3440,11 @@
    "type" : "bool",
    "comment" : "(boolean, default false) Sparse update",
    "generated" : 0
+ }, { 
+   "name" : "padding_idx",
+   "type" : "long",
+   "comment" : "(int64, default -1) If the value is -1, it makes no effect to lookup. Otherwise the given value indicates padding the output with zeros whenever lookup encounters it in Ids.",
+   "generated" : 0
  } ] 
 },{
  "type" : "lod_tensor_to_array",

develop/doc_cn/api/v2/fluid/layers.html

@@ -328,10 +328,11 @@ layer.</li>
 <h2>embedding<a class="headerlink" href="#embedding" title="永久链接至标题">¶</a></h2>
 <dl class="function">
 <dt>
-<code class="descclassname">paddle.v2.fluid.layers.</code><code class="descname">embedding</code><span class="sig-paren">(</span><em>input</em>, <em>size</em>, <em>is_sparse=False</em>, <em>param_attr=None</em>, <em>dtype='float32'</em><span class="sig-paren">)</span></dt>
+<code class="descclassname">paddle.v2.fluid.layers.</code><code class="descname">embedding</code><span class="sig-paren">(</span><em>input</em>, <em>size</em>, <em>is_sparse=False</em>, <em>padding_idx=None</em>, <em>param_attr=None</em>, <em>dtype='float32'</em><span class="sig-paren">)</span></dt>
 <dd><p><strong>Embedding Layer</strong></p>
-<p>This layer is used to lookup a vector of IDs, provided by <em>input</em>, in a lookup table.
-The result of this lookup is the embedding of each ID in the <em>input</em>.</p>
+<p>This layer is used to lookup embeddings of IDs, provided by <code class="xref py py-attr docutils literal"><span class="pre">input</span></code>, in
+a lookup table. The result of this lookup is the embedding of each ID in the
+<code class="xref py py-attr docutils literal"><span class="pre">input</span></code>.</p>
 <p>All the input variables are passed in as local variables to the LayerHelper
 constructor.</p>
 <table class="docutils field-list" frame="void" rules="none">
@@ -339,9 +340,16 @@ constructor.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>Variable</em>) &#8211; Input to the function</li>
-<li><strong>size</strong> (<em>tuple|list|None</em>) &#8211; Shape of the look up table parameter</li>
-<li><strong>is_sparse</strong> (<em>bool</em>) &#8211; Boolean flag that specifying whether the input is sparse</li>
+<li><strong>input</strong> (<em>Variable</em>) &#8211; The tensor variable containing the IDs.</li>
+<li><strong>size</strong> (<em>tuple|list</em>) &#8211; The shape of the look up table parameter. It should
+have two elements which indicate the size of the dictionary of
+embeddings and the size of each embedding vector respectively.</li>
+<li><strong>is_sparse</strong> (<em>bool</em>) &#8211; The flag indicating whether to use sparse update.</li>
+<li><strong>padding_idx</strong> (<em>int|long|None</em>) &#8211; If <code class="xref py py-attr docutils literal"><span class="pre">None</span></code>, it makes no effect to lookup.
+Otherwise the given <code class="xref py py-attr docutils literal"><span class="pre">padding_idx</span></code> indicates padding the output
+with zeros whenever lookup encounters it in <code class="xref py py-attr docutils literal"><span class="pre">input</span></code>. If
+<span class="math">\(padding_idx &lt; 0\)</span>, the padding_idx to use in lookup is
+<span class="math">\(size[0] + dim\)</span>.</li>
 <li><strong>param_attr</strong> (<em>ParamAttr</em>) &#8211; Parameters for this layer</li>
 <li><strong>dtype</strong> (<em>np.dtype|core.DataType|str</em>) &#8211; The type of data : float32, float_16, int etc</li>
 </ul>
@@ -1063,10 +1071,11 @@ that need to be summed up.</td>
 <h2>assign<a class="headerlink" href="#assign" title="永久链接至标题">¶</a></h2>
 <dl class="function">
 <dt>
-<code class="descclassname">paddle.v2.fluid.layers.</code><code class="descname">embedding</code><span class="sig-paren">(</span><em>input</em>, <em>size</em>, <em>is_sparse=False</em>, <em>param_attr=None</em>, <em>dtype='float32'</em><span class="sig-paren">)</span></dt>
+<code class="descclassname">paddle.v2.fluid.layers.</code><code class="descname">embedding</code><span class="sig-paren">(</span><em>input</em>, <em>size</em>, <em>is_sparse=False</em>, <em>padding_idx=None</em>, <em>param_attr=None</em>, <em>dtype='float32'</em><span class="sig-paren">)</span></dt>
 <dd><p><strong>Embedding Layer</strong></p>
-<p>This layer is used to lookup a vector of IDs, provided by <em>input</em>, in a lookup table.
-The result of this lookup is the embedding of each ID in the <em>input</em>.</p>
+<p>This layer is used to lookup embeddings of IDs, provided by <code class="xref py py-attr docutils literal"><span class="pre">input</span></code>, in
+a lookup table. The result of this lookup is the embedding of each ID in the
+<code class="xref py py-attr docutils literal"><span class="pre">input</span></code>.</p>
 <p>All the input variables are passed in as local variables to the LayerHelper
 constructor.</p>
 <table class="docutils field-list" frame="void" rules="none">
@@ -1074,9 +1083,16 @@ constructor.</p>
 <col class="field-body" />
 <tbody valign="top">
 <tr class="field-odd field"><th class="field-name">参数:</th><td class="field-body"><ul class="first simple">
-<li><strong>input</strong> (<em>Variable</em>) &#8211; Input to the function</li>
-<li><strong>size</strong> (<em>tuple|list|None</em>) &#8211; Shape of the look up table parameter</li>
-<li><strong>is_sparse</strong> (<em>bool</em>) &#8211; Boolean flag that specifying whether the input is sparse</li>
+<li><strong>input</strong> (<em>Variable</em>) &#8211; The tensor variable containing the IDs.</li>
+<li><strong>size</strong> (<em>tuple|list</em>) &#8211; The shape of the look up table parameter. It should
+have two elements which indicate the size of the dictionary of
+embeddings and the size of each embedding vector respectively.</li>
+<li><strong>is_sparse</strong> (<em>bool</em>) &#8211; The flag indicating whether to use sparse update.</li>
+<li><strong>padding_idx</strong> (<em>int|long|None</em>) &#8211; If <code class="xref py py-attr docutils literal"><span class="pre">None</span></code>, it makes no effect to lookup.
+Otherwise the given <code class="xref py py-attr docutils literal"><span class="pre">padding_idx</span></code> indicates padding the output
+with zeros whenever lookup encounters it in <code class="xref py py-attr docutils literal"><span class="pre">input</span></code>. If
+<span class="math">\(padding_idx &lt; 0\)</span>, the padding_idx to use in lookup is
+<span class="math">\(size[0] + dim\)</span>.</li>
 <li><strong>param_attr</strong> (<em>ParamAttr</em>) &#8211; Parameters for this layer</li>
 <li><strong>dtype</strong> (<em>np.dtype|core.DataType|str</em>) &#8211; The type of data : float32, float_16, int etc</li>
 </ul>

develop/doc_cn/api/v2/fluid/nets.html

@@ -249,7 +249,7 @@
 <h2>img_conv_group<a class="headerlink" href="#img-conv-group" title="永久链接至标题">¶</a></h2>
 <dl class="function">
 <dt>
-<code class="descclassname">paddle.v2.fluid.nets.</code><code class="descname">img_conv_group</code><span class="sig-paren">(</span><em>input</em>, <em>conv_num_filter</em>, <em>pool_size</em>, <em>conv_padding=1</em>, <em>conv_filter_size=3</em>, <em>conv_act=None</em>, <em>param_attr=None</em>, <em>conv_with_batchnorm=False</em>, <em>conv_batchnorm_drop_rate=None</em>, <em>pool_stride=1</em>, <em>pool_type=None</em>, <em>use_cudnn=True</em><span class="sig-paren">)</span></dt>
+<code class="descclassname">paddle.v2.fluid.nets.</code><code class="descname">img_conv_group</code><span class="sig-paren">(</span><em>input</em>, <em>conv_num_filter</em>, <em>pool_size</em>, <em>conv_padding=1</em>, <em>conv_filter_size=3</em>, <em>conv_act=None</em>, <em>param_attr=None</em>, <em>conv_with_batchnorm=False</em>, <em>conv_batchnorm_drop_rate=0.0</em>, <em>pool_stride=1</em>, <em>pool_type=None</em>, <em>use_cudnn=True</em><span class="sig-paren">)</span></dt>
 <dd><p>Image Convolution Group, Used for vgg net.</p>
 </dd></dl>