<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"><htmlxmlns="http://www.w3.org/1999/xhtml"><head><metahttp-equiv="Content-Type"content="text/html; charset=UTF-8"/><title>TensorFlow.GenOps.Core</title><linkhref="ocean.css"rel="stylesheet"type="text/css"title="Ocean"/><scriptsrc="haddock-util.js"type="text/javascript"></script><scripttype="text/javascript">//<![CDATA[
window.onload = function () {pageLoad();setSynopsis("mini_TensorFlow-GenOps-Core.html");};
</script></head><body><divid="package-header"><ulclass="links"id="page-menu"><li><ahref="src/TensorFlow-GenOps-Core.html">Source</a></li><li><ahref="index.html">Contents</a></li><li><ahref="doc-index.html">Index</a></li></ul><pclass="caption">tensorflow-core-ops-0.1.0.0: Haskell wrappers for Core Tensorflow Ops.</p></div><divid="content"><divid="module-header"><tableclass="info"><tr><th>Safe Haskell</th><td>None</td></tr><tr><th>Language</th><td>Haskell2010</td></tr></table><pclass="caption">TensorFlow.GenOps.Core</p></div><divid="synopsis"><pid="control.syn"class="caption expander"onclick="toggleSection('syn')">Synopsis</p><ulid="section.syn"class="hide"onclick="toggleSection('syn')"><liclass="src short"><ahref="#v:_HostRecv">_HostRecv</a> :: <spanclass="keyword">forall</span> tensor_type. <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tensor_type =><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a> -><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> tensor_type)</li><liclass="src short"><ahref="#v:_HostSend">_HostSend</a> :: <spanclass="keyword">forall</span> v1 t. <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t =><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a> -><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t -><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></li><liclass="src short"><ahref="#v:_Recv">_Recv</a> :: <spanclass="keyword">forall</span> tensor_type. <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tensor_type =><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a> -><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> tensor_type)</li><liclass="src short"><ahref="#v:_Send">_Send</a> :: <spanclass="keyword">forall</span> v1 t. <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t =><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a> -><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t -><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></li><liclass="src short"><ahref="#v:noOp">noOp</a> :: <ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></li><liclass="src short"><ahref="#v:_Retval">_Retval</a> :: <spanclass="keyword">forall</span> v1 t. <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t =><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a> -><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t -><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></li><liclass="src short"><ahref="#v:_Arg">_Arg</a> :: <spanclass="keyword">forall</span> t. <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t =><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a> -><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</li><liclass="src short"><ahref="#v:quantizedBatchNormWithGlobalNormalization">quantizedBatchNormWithGlobalNormalization</a> :: <spanclass="keyword">forall</span> v1 v2 v3 v4 v5 v6 v7 v8 v9 v10 v11 v12 v13 v14 v15 tinput out_type. (<ahref="../tenso
input on device memory.</p></div></div><divclass="top"><pclass="src"><aname="v:_HostSend"class="def">_HostSend</a><ahref="src/TensorFlow-GenOps-Core.html#_HostSend"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>send_device_incarnation</strong>: The current incarnation of send_device.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>tensor</strong>: The tensor to send.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Sends the named tensor from send_device to recv_device.</p><p>_HostSend requires its input on host memory whereas _Send requires its
input on device memory.</p></div></div><divclass="top"><pclass="src"><aname="v:_Recv"class="def">_Recv</a><ahref="src/TensorFlow-GenOps-Core.html#_Recv"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tensor_type</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>send_device_incarnation</strong>: The current incarnation of send_device.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> tensor_type)</td><tdclass="doc"><p><strong>tensor</strong>: The tensor to receive.</p></td></tr></table></div><divclass="doc"><p>Receives the named tensor from send_device on recv_device.</p></div></div><divclass="top"><pclass="src"><aname="v:_Send"class="def">_Send</a><ahref="src/TensorFlow-GenOps-Core.html#_Send"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>send_device_incarnation</strong>: The current incarnation of send_device.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>tensor</strong>: The tensor to send.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Sends the named tensor from send_device to recv_device.</p></div></div><divclass="top"><pclass="src"><aname="v:noOp"class="def">noOp</a> :: <ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a><ahref="src/TensorFlow-GenOps-Core.html#noOp"class="link">Source</a></p><divclass="doc"><p>Does nothing. Only useful as a placeholder for control edges.</p></div></div><divclass="top"><pclass="src"><aname="v:_Retval"class="def">_Retval</a><ahref="src/TensorFlow-GenOps-Core.html#_Retval"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>index</strong>: This return value is the index-th return value of the function.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: The return value.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>A graph node which represents a return value of a function.</p></div></div><divclass="top"><pclass="src"><aname="v:_Arg"class="def">_Arg</a><ahref="src/TensorFlow-GenOps-Core.html#_Arg"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>index</strong>: This argument is the index-th a
needs to be multiplied with gamma.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>variance_epsilon</strong>: A small float number to avoid dividing by 0.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 tinput</td><tdclass="doc"><p><strong>t</strong>: A 4D input Tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>t_min</strong>: The value represented by the lowest quantized input.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>t_max</strong>: The value represented by the highest quantized input.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 tinput</td><tdclass="doc"><p><strong>m</strong>: A 1D mean Tensor with size matching the last dimension of t.
This is the first output from tf.nn.moments,
or a saved moving average thereof.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>m_min</strong>: The value represented by the lowest quantized mean.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>m_max</strong>: The value represented by the highest quantized mean.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v7 tinput</td><tdclass="doc"><p><strong>v</strong>: A 1D variance Tensor with size matching the last dimension of t.
This is the second output from tf.nn.moments,
or a saved moving average thereof.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v8 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>v_min</strong>: The value represented by the lowest quantized variance.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v9 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>v_max</strong>: The value represented by the highest quantized variance.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v10 tinput</td><tdclass="doc"><p><strong>beta</strong>: A 1D beta Tensor with size matching the last dimension of t.
An offset to be added to the normalized tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v11 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>beta_min</strong>: The value represented by the lowest quantized offset.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v12 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>beta_max</strong>: The value represented by the highest quantized offset.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v13 tinput</td><tdclass="doc"><p><strong>gamma</strong>: A 1D gamma Tensor with size matching the last dimension of t.
If "scale_after_normalization" is true, this tensor will be multiplied
with the normalized tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v14 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>gamma_min</strong>: The value represented by the lowest quantized gamma.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v15 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>gamma_max</strong>: The value represented by the highest quantized gamma.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>result</strong>, <strong>result_min</strong>, <strong>result_max</strong>)</p><ul><li><strong>result</strong></li><li><strong>result_min</strong></li><li><strong>result_max</strong></li></ul></td></tr></table></div><divclass="doc"><p>Quantized Batch normalization.</p><p>This op is deprecated and will be removed in the future. Prefer
`tf.nn.batch_normalization`.</p></div></div><divclass="top"><pclass="src"><aname="v:quantizedRelu6"class="def">quantizedRelu6</a><ahref="src/TensorFlow-GenOps-Core.html#quantizedRelu6"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tinput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` tinput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` out_type)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 tinput</td><tdclass="doc"><p><strong>features</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>min_features</strong>: The float value that the lowest quantized value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>max_features</strong>: The float value that the highest quantized value represents.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>activations</strong>, <strong>min_activations</strong>, <strong>max_activations</strong>)</p><ul><li><strong>activations</strong>: Has the same output shape as "features".</li><li><strong>min_activations</strong>: The float value that the lowest quantized value represents.</li><li><strong>max_activations</strong>: The float value that the highest quantized value represents.</li></ul></td></tr></table></div><divclass="doc"><p>Computes Quantized Rectified Linear 6: `min(max(features, 0), 6)`</p></div></div><divclass="top"><pclass="src"><aname="v:quantizedBiasAdd"class="def">quantizedBiasAdd</a><ahref="src/TensorFlow-GenOps-Core.html#quantizedBiasAdd"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t1, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` t1, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t2, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a
w.r.t. the output of <code>fractional_avg_pool</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>row_pooling_sequence</strong>: row pooling sequence, form pooling region with
col_pooling_sequence.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>col_pooling_sequence</strong>: column pooling sequence, form pooling region with
row_pooling sequence.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 4-D. Gradients w.r.t. the input of <code>fractional_avg_pool</code>.</p></td></tr></table></div><divclass="doc"><p>Computes gradient of the FractionalAvgPool function.</p><p>Unlike FractionalMaxPoolGrad, we don't need to find arg_max for
FractionalAvgPoolGrad, we just need to evenly back-propagate each element of
out_backprop to those indices that form the same pooling cell. Therefore, we
just need to know the shape of original input tensor, instead of the whole
tensor.</p></div></div><divclass="top"><pclass="src"><aname="v:fractionalMaxPoolGrad"class="def">fractionalMaxPoolGrad</a><ahref="src/TensorFlow-GenOps-Core.html#fractionalMaxPoolGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>orig_input</strong>: Original input for <code>fractional_max_pool</code></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>orig_output</strong>: Original output for <code>fractional_max_pool</code></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>out_backprop</strong>: 4-D with shape `[batch, height, width, channels]`. Gradients
w.r.t. the output of <code>fractional_max_pool</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>row_pooling_sequence</strong>: row pooling sequence, form pooling region with
col_pooling_sequence.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>col_pooling_sequence</strong>: column pooling sequence, form pooling region with
row_pooling sequence.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 4-D. Gradients w.r.t. the input of <code>fractional_max_pool</code>.</p></td></tr></table></div><divclass="doc"><p>Computes gradient of the FractionalMaxPool function.</p></div></div><divclass="top"><pclass="src"><aname="v:fractionalMaxPool"class="def">fractionalMaxPool</a><ahref="src/TensorFlow-GenOps-Core.html#fractionalMaxPool"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>value</strong>: 4-D with shape `[batch, height, width, channels]`.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p>(<strong>output</strong>, <strong>row_pooling_sequence</strong>, <strong>col_pooling_sequence</strong>)</p><ul><li><strong>output</strong>: output tensor after fractional max pooling.</li><li><strong>row_pooling_sequence</strong>: row pooling sequence, needed to calculate gradient.</li><li><strong>col_pooling_sequence</strong>: column pooling sequence, needed to calculate gradient.</li></ul></td></tr></table></div><divclass="doc"><p>Performs fractional max pooling on the input.</p><p>Fractional max pooling is slightly different than regular max pooling. In
regular max pooling, you downsize an input set by taking the maximum value of
smaller N x N subsections of the set (often 2x2), and try to reduce the set by
a factor of N, where N is an integer. Fractional max pooling, as you might
expect from the word "fractional", means that the overall reduction ratio N
does not have to be an integer.</p><p>The sizes of the pooling regions are generated randomly but are fairly uniform.
For example, let's look at the height dimension, and the constraints on the
list of rows that will be pool boundaries.</p><p>First we define the following:</p><ol><li>input_row_length : the number of rows from the input set</li><li>output_row_length : which will be smaller than the input</li><li>alpha = input_row_length / output_row_length : our reduction ratio</li><li>K = floor(alpha)</li><li>row_pooling_sequence : this is the result list of pool boundary rows</li></ol><p>Then, row_pooling_sequence should satisfy:</p><ol><li>a[0] = 0 : the first value of the sequence is 0</li><li>a[end] = input_row_length : the last value of the sequence is the size</li><li>K <= (a[i+1] - a[i]) <= K+1 : all intervals are K or K+1 size</li><li>length(row_pooling_sequence) = output_row_length+1</li></ol><p>For more details on fractional max pooling, see this paper:
<ahref="http://arxiv.org/abs/1412.6071">Benjamin Graham, Fractional Max-Pooling</a></p></div></div><divclass="top"><pclass="src"><aname="v:topK"class="def">topK</a><ahref="src/TensorFlow-GenOps-Core.html#topK"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>k</strong>: Number of top elements to look for along the last dimension (along each
row for matrices).</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: 1-D or higher with last dimension at least <code>k</code>.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>)</td><tdclass="doc"><p>(<strong>values</strong>, <strong>indices</strong>)</p><ul><li><strong>values</strong>: The <code>k</code> largest elements along each last dimensional slice.</li><li><strong>indices</strong>: The indices of <code>values</code> within the last dimension of <code>input</code>.</li></ul></td></tr></table></div><divclass="doc"><p>Finds values and indices of the <code>k</code> largest elements for the last dimension.</p><p>If the input is a vector (rank-1), finds the <code>k</code> largest entries in the vector
and outputs their values and indices as vectors. Thus `values[j]` is the
<code>j</code>-th largest entry in <code>input</code>, and its index is `indices[j]`.</p><p>For matrices (resp. higher rank input), computes the top <code>k</code> entries in each
row (resp. vector along the last dimension). Thus,</p><p>values.shape = indices.shape = input.shape[:-1] + [k]</p><p>If two elements are equal, the lower-index element appears first.</p><p>If <code>k</code> varies dynamically, use <code>TopKV2</code> below.</p></div></div><divclass="top"><pclass="src"><aname="v:inTopK"class="def">inTopK</a><ahref="src/TensorFlow-GenOps-Core.html#inTopK"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>k</strong>: Number of top elements to look at for computing precision.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>predictions</strong>: A <code>batch_size</code> x <code>classes</code> tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>targets</strong>: A <code>batch_size</code> vector of class ids.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>precision</strong>: Computed Precision at <code>k</code> as a `bool Tensor`.</p></td></tr></table></div><divclass="doc"><p>Says whether the targets are in the top <code>K</code> predictions.</p><p>This outputs a <code>batch_size</code> bool array, an entry `out[i]` is <code>true</code> if the
prediction for the target class is among the top <code>k</code> predictions among
all predictions for example <code>i</code>. Note that the behavior of <code>InTopK</code> differs
from the <code>TopK</code> op in its handling of ties; if multiple classes have the
same prediction value and straddle the top-<code>k</code> boundary, all of those
classes are considered to be in the top <code>k</code>.</p><p>More formally, let</p><p>\(predictions_i\) be the predictions for all classes for example <code>i</code>,
\(targets_i\) be the target class for example <code>i</code>,
\(out_i\) be the output for example <code>i</code>,</p><p>$$out_i = predictions_{i, targets_i} in TopKIncludingTies(predictions_i)$$</p></div></div><divclass="top"><pclass="src"><aname="v:sparseSoftmaxCrossEntropyWithLogits"class="def">sparseSoftmaxCrossEntropyWithLogits</a><ahref="src/TensorFlow-GenOps-Core.html#sparseSoftmaxCrossEntropyWithLogits"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tlabels, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tlabels)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>features</strong>: batch_size x num_classes matrix</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tlabels</td><tdclass="doc"><p><strong>labels</strong>: batch_size vector with values in [0, num_classes).
This is the label for the given minibatch entry.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p>(<strong>loss</strong>, <strong>backprop</strong>)</p><ul><li><strong>loss</strong>: Per example loss (batch_size vector).</li><li><strong>backprop</strong>: backpropagated gradients (batch_size x num_classes matrix).</li></ul></td></tr></table></div><divclass="doc"><p>Computes softmax cross entropy cost and gradients to backpropagate.</p><p>Unlike <code>SoftmaxCrossEntropyWithLogits</code>, this operation does not accept
a matrix of label probabilities, but rather a single label per row
of features. This label is considered to have probability 1.0 for the
given row.</p><p>Inputs are the logits, not probabilities.</p></div></div><divclass="top"><pclass="src"><aname="v:softmaxCrossEntropyWithLogits"class="def">softmaxCrossEntropyWithLogits</a><ahref="src/TensorFlow-GenOps-Core.html#softmaxCrossEntropyWithLogits"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>features</strong>: batch_size x num_classes matrix</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>labels</strong>: batch_size x num_classes matrix
The caller must ensure that each batch of labels represents a valid
probability distribution.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p>(<strong>loss</strong>, <strong>backprop</strong>)</p><ul><li><strong>loss</strong>: Per example loss (batch_size vector).</li><li><strong>backprop</strong>: backpropagated gradients (batch_size x num_classes matrix).</li></ul></td></tr></table></div><divclass="doc"><p>Computes softmax cross entropy cost and gradients to backpropagate.</p><p>Inputs are the logits, not probabilities.</p></div></div><divclass="top"><pclass="src"><aname="v:logSoftmax"class="def">logSoftmax</a><ahref="src/TensorFlow-GenOps-Core.html#logSoftmax"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>logits</strong>: 2-D with shape `[batch_size, num_classes]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>logsoftmax</strong>: Same shape as <code>logits</code>.</p></td></tr></table></div><divclass="doc"><p>Computes log softmax activations.</p><p>For each batch <code>i</code> and class <code>j</code> we have</p><p>logsoftmax[i, j] = logits[i, j] - log(sum(exp(logits[i])))</p></div></div><divclass="top"><pclass="src"><aname="v:softsignGrad"class="def">softsignGrad</a><ahref="src/TensorFlow-GenOps-Core.html#softsignGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>gradients</strong>: The backpropagated gradients to the corresponding softsign operation.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>features</strong>: The features passed as input to the corresponding softsign operation.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>backprops</strong>: The gradients: `gradients / (1 + abs(-features)) ** 2`.</p></td></tr></table></div><divclass="doc"><p>Computes softsign gradients for a softsign operation.</p></div></div><divclass="top"><pclass="src"><aname="v:softplus"class="def">softplus</a><ahref="src/TensorFlow-GenOps-Core.html#softplus"class="link">S
<code>gradients</code> otherwise.</p></td></tr></table></div><divclass="doc"><p>Computes gradients for the exponential linear (Elu) operation.</p></div></div><divclass="top"><pclass="src"><aname="v:elu"class="def">elu</a><ahref="src/TensorFlow-GenOps-Core.html#elu"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>features</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>activations</strong></p></td></tr></table></div><divclass="doc"><p>Computes exponential linear: `exp(features) - 1` if < 0, <code>features</code> otherwise.</p><p>See <ahref="http://arxiv.org/abs/1511.07289">Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs)</a></p></div></div><divclass="top"><pclass="src"><aname="v:relu6"class="def">relu6</a><ahref="src/TensorFlow-GenOps-Core.html#relu6"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>features</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>activations</strong></p></td></tr></table></div><divclass="doc"><p>Computes rectified linear 6: `min(max(features, 0), 6)`.</p></div></div><divclass="top"><pclass="src"><aname="v:reluGrad"class="def">reluGrad</a><ahref="src/TensorFlow-GenOps-Core.html#reluGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty">
the outputs of that operation (both work equivalently).</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>backprops</strong>: `gradients * (features > 0)`.</p></td></tr></table></div><divclass="doc"><p>Computes rectified linear gradients for a Relu operation.</p></div></div><divclass="top"><pclass="src"><aname="v:dilation2DBackpropInput"class="def">dilation2DBackpropInput</a><ahref="src/TensorFlow-GenOps-Core.html#dilation2DBackpropInput"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: 4-D with shape `[batch, in_height, in_width, depth]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>filter</strong>: 3-D with shape `[filter_height, filter_width, depth]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>out_backprop</strong>: 4-D with shape `[batch, out_height, out_width, depth]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>in_backprop</strong>: 4-D with shape `[batch, in_height, in_width, depth]`.</p></td></tr></table></div><divclass="doc"><p>Computes the gradient of morphological 2-D dilation with respect to the input.</p></div></div><divclass="top"><pclass="src"><aname="v:maxPoolGrad"class="def">maxPoolGrad</a><ahref="src/TensorFlow-GenOps-Core.html#maxPoolGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>orig_input</strong>: The original input tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>orig_output</strong>: The original output tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>grad</strong>: 4-D. Gradients w.r.t. the output of <code>max_pool</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Gradients w.r.t. the input to <code>max_pool</code>.</p></td></tr></table></div><divclass="doc"><p>Computes gradients of the maxpooling function.</p><
where <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code> is a 5-D
tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>out_backprop</strong>: Backprop signal of shape `[batch, out_depth, out_rows, out_cols,
out_channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Computes the gradients of 3-D convolution with respect to the filter.</p></div></div><divclass="top"><pclass="src"><aname="v:conv3DBackpropFilter"class="def">conv3DBackpropFilter</a><ahref="src/TensorFlow-GenOps-Core.html#conv3DBackpropFilter"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Shape `[batch, depth, rows, cols, in_channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>filter</strong>: Shape `[depth, rows, cols, in_channels, out_channels]`.
<code>in_channels</code> must match between <code>input</code> and <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>out_backprop</strong>: Backprop signal of shape `[batch, out_depth, out_rows, out_cols,
out_channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Computes the gradients of 3-D convolution with respect to the filter.</p></div></div><divclass="top"><pclass="src"><aname="v:conv3D"class="def">conv3D</a><ahref="src/TensorFlow-GenOps-Core.html#conv3D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Shape `[batch, in_depth, in_height, in_width, in_channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>filter</strong>: Shape `[filter_depth, filter_height, filter_width, in_channels,
out_channels]`. <code>in_channels</code> must match between <code>input</code> and <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Computes a 3-D convolution given 5-D <code>input</code> and <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code> tensors.</p><p>In signal processing, cross-correlation is a measure of similarity of
two waveforms as a function of a time-lag applied to one of them. This
is also known as a sliding dot product or sliding inner-product.</p><p>Our Conv3D implements a form of cross-correlation.</p></div></div><divclass="top"><pclass="src"><aname="v:depthwiseConv2dNativeBackpropFilter"class="def">depthwiseConv2dNativeBackpropFilter</a><ahref="src/TensorFlow-GenOps-Core.html#depthwiseConv2dNativeBackpropFilter"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: 4-D with shape `[batch, in_height, in_width, in_channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>filter_sizes</strong>: An integer vector representing the tensor shape of <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code>,
where <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code> is a 4-D
Gradients w.r.t. the output of the convolution.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 4-D with shape
the <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code> input of the convolution.</p></td></tr></table></div><divclass="doc"><p>Computes the gradients of depthwise convolution with respect to the filter.</p></div></div><divclass="top"><pclass="src"><aname="v:conv2DBackpropFilter"class="def">conv2DBackpropFilter</a><ahref="src/TensorFlow-GenOps-Core.html#conv2DBackpropFilter"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: 4-D with shape `[batch, in_height, in_width, in_channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>filter_sizes</strong>: An integer vector representing the tensor shape of <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code>,
where <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code> is a 4-D
Gradients w.r.t. the output of the convolution.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 4-D with shape
the <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code> input of the convolution.</p></td></tr></table></div><divclass="doc"><p>Computes the gradients of convolution with respect to the filter.</p></div></div><divclass="top"><pclass="src"><aname="v:conv2DBackpropInput"class="def">conv2DBackpropInput</a><ahref="src/TensorFlow-GenOps-Core.html#conv2DBackpropInput"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>input_sizes</strong>: An integer vector representing the shape of <code>input</code>,
where <code>input</code> is a 4-D `[batch, height, width, channels]` tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>filter</strong>: 4-D with shape
Gradients w.r.t. the output of the convolution.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 4-D with shape `[batch, in_height, in_width, in_channels]`. Gradient
w.r.t. the input of the convolution.</p></td></tr></table></div><divclass="doc"><p>Computes the gradients of convolution with respect to the input.</p></div></div><divclass="top"><pclass="src"><aname="v:conv2D"class="def">conv2D</a><ahref="src/TensorFlow-GenOps-Core.html#conv2D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>filter</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Computes a 2-D convolution given 4-D <code>input</code> and <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code> tensors.</p><p>Given an input tensor of shape `[batch, in_height, in_width, in_channels]`
and a filter / kernel tensor of shape
`[filter_height, filter_width, in_channels, out_channels]`, this op
performs the following:</p><ol><li>Flattens the filter to a 2-D matrix with shape
`[filter_height * filter_width * in_channels, output_channels]`.</li><li>Extracts image patches from the input tensor to form a *virtual*
tensor of shape `[batch, out_height, out_width,
filter_height * filter_width * in_channels]`.</li><li>For each patch, right-multiplies the filter matrix and the image patch
vector.</li></ol><p>In detail, with the default NHWC format,</p><p>output[b, i, j, k] =
filter[di, dj, q, k]</p><p>Must have `strides[0] = strides[3] = 1`. For the most common case of the same
horizontal and vertices strides, `strides = [1, stride, stride, 1]`.</p></div></div><divclass="top"><pclass="src"><aname="v:biasAdd"class="def">biasAdd</a><ahref="src/TensorFlow-GenOps-Core.html#biasAdd"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>value</strong>: Any number of dimensions.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>bias</strong>: 1-D with size the last dimension of <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Broadcasted sum of <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code> and <code>bias</code>.</p></td></tr></table></div><divclass="doc"><p>Adds <code>bias</code> to <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>.</p><p>This is a special case of `tf.add` where <code>bias</code> is restricted to be 1-D.
Broadcasting is supported, so <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code> may have any number of dimensions.</p></div></div><divclass="top"><pclass="src"><aname="v:fusedBatchNorm"class="def">fusedBatchNorm</a><ahref="src/TensorFlow-GenOps-Core.html#fusedBatchNorm"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong>: A 4D Tensor for input data.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>scale</strong>: A 1D Tensor for scaling factor, to scale the normalized x.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>offset</strong>: A 1D Tensor for offset, to shift to the normalized x.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>mean</strong>: A 1D Tensor for population mean. Used for inference only;
must be empty for training.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>variance</strong>: A 1D Tensor for population variance. Used for inference only;
must be empty for training.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p>(<strong>y</strong>, <strong>batch_mean</strong>, <strong>batch_variance</strong>, <strong>reserve_space_1</strong>, <strong>reserve_space_2</strong>)</p><ul><li><strong>y</strong>: A 4D Tensor for output data.</li><li><strong>batch_mean</strong>: A 1D Tensor for the computed batch mean, to be used by TensorFlow
to compute the running mean.</li><li><strong>batch_variance</strong>: A 1D Tensor for the computed batch variance, to be used by
TensorFlow to compute the running variance.</li><li><strong>reserve_space_1</strong>: A 1D Tensor for the computed batch mean, to be reused
in the gradient computation.</li><li><strong>reserve_space_2</strong>: A 1D Tensor for the computed batch variance (inverted variance
in the cuDNN case), to be used in the gradient computation.</li></ul></td></tr></table></div><divclass="doc"><p>Batch normalization.</p><p>Note that the size of 4D Tensors are defined by either <ahref="NHWC.html">NHWC</a> or <ahref="NCHW.html">NCHW</a>.
The size of 1D Tensors matches the dimension C of the 4D Tensors.</p></div></div><divclass="top"><pclass="src"><aname="v:batchNormWithGlobalNormalizationGrad"class="def">batchNormWithGlobalNormalizationGrad</a><ahref="src/TensorFlow-GenOps-Core.html#batchNormWithGlobalNormalizationGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>scale_after_normalization</strong>: A bool indicating whether the resulted tensor
needs to be multiplied with gamma.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>variance_epsilon</strong>: A small float number to avoid dividing by 0.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>t</strong>: A 4D input Tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>m</strong>: A 1D mean Tensor with size matching the last dimension of t.
This is the first output from tf.nn.moments,
or a saved moving average thereof.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>v</strong>: A 1D variance Tensor with size matching the last dimension of t.
This is the second output from tf.nn.moments,
or a saved moving average thereof.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>gamma</strong>: A 1D gamma Tensor with size matching the last dimension of t.
If "scale_after_normalization" is true, this Tensor will be multiplied
with the normalized Tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>backprop</strong>: 4D backprop Tensor.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p>(<strong>dx</strong>, <strong>dm</strong>, <strong>dv</strong>, <strong>db</strong>, <strong>dg</strong>)</p><ul><li><strong>dx</strong>: 4D backprop tensor for input.</li><li><strong>dm</strong>: 1D backprop tensor for mean.</li><li><strong>dv</strong>: 1D backprop tensor for variance.</li><li><strong>db</strong>: 1D backprop tensor for beta.</li><li><strong>dg</strong>: 1D backprop tensor for gamma.</li></ul></td></tr></table></div><divclass="doc"><p>Gradients for batch normalization.</p><p>This op is deprecated. See `tf.nn.batch_normalization`.</p></div></div><divclass="top"><pclass="src"><aname="v:batchFFT3D"class="def">batchFFT3D</a><ahref="src/TensorFlow-GenOps-Core.html#batchFFT3D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:batchIFFT2D"class="def">batchIFFT2D</a><ahref="src/TensorFlow-GenOps-Core.html#batchIFFT2D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:avgPool"class="def">avgPool</a><ahref="src/TensorFlow-GenOps-Core.html#avgPool"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Ten
window in <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>.</p></div></div><divclass="top"><pclass="src"><aname="v:batchFFT2D"class="def">batchFFT2D</a><ahref="src/TensorFlow-GenOps-Core.html#batchFFT2D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:batchFFT"class="def">batchFFT</a><ahref="src/TensorFlow-GenOps-Core.html#batchFFT"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:requantizationRange"class="def">requantizationRange</a><ahref="src/TensorFlow-GenOps-Core.html#requantizationRange"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tinput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` tinput)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 tinput</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>input_min</strong>: The float value that the minimum quantized input value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>input_max</strong>: The float value that the maximum quantized input value represents.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>output_min</strong>, <strong>output_max</strong>)</p><ul><li><strong>output_min</strong>: The computed min output.</li><li><strong>output_max</strong>: the computed max output.</li></ul></td></tr></table></div><divclass="doc"><p>Given a quan
typically used to produce the requested_output_min and requested_output_max for
Requantize.</p></div></div><divclass="top"><pclass="src"><aname="v:requantize"class="def">requantize</a><ahref="src/TensorFlow-GenOps-Core.html#requantize"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tinput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` tinput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` out_type)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 tinput</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>input_min</strong>: The float value that the minimum quantized input value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>input_max</strong>: The float value that the maximum quantized input value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>requested_output_min</strong>: The float value that the minimum quantized output value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>requested_output_max</strong>: The float value that the maximum quantized output value represents.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>output</strong>, <strong>output_min</strong>, <strong>output_max</strong>)</p><ul><li><strong>output</strong></li><li><strong>output_min</strong>: The requested_output_min value is copied into this output.</li><li><strong>output_max</strong>: The requested_output_max value is copied into this output.</li></ul></td></tr></table></div><divclass="doc"><p>Convert the quantized <code>input</code> tensor into a lower-precision <code>output</code>, using the</p><p>output range specified with <code>requested_output_min</code> and <code>requested_output_max</code>.</p><dl><dt>input_min, input_max</dt><dd>are scalar floats that specify the range for the float
interpretation of the <code>input</code> data. For example, if input_min is -1.0f and
input_max is 1.0f, and we are dealing with quint16 quantized data, then a 0
value in the 16-bit data should be interpreted as -1.0f, and a 65535 means 1.0f.</dd></dl></div></div><divclass="top"><pclass="src"><aname="v:quantizeDownAndShrinkRange"class="def">quantizeDownAndShrinkRange</a><ahref="src/TensorFlow-GenOps-Core.html#quantizeDownAndShrinkRange"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tinput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` tinput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` out_type)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 tinput</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>input_min</strong>: The float value that the minimum quantized input value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>input_max</strong>: The float value that the maximum quantized input value represents.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>output</strong>, <strong>output_min</strong>, <strong>output_max</strong>)</p><ul><li><strong>output</strong></li><li><strong>output_min</strong>: The float value that the minimum quantized output value represents.</li><li><strong>output_max</strong>: The float value that the maximum quantized output value represents.</li></ul></td></tr></table></div><divclass="doc"><p>Convert the quantized <code>input</code> tensor into a lower-precision <code>output</code>, using the</p><p>actual distribution of the values to maximize the usage of the lower bit depth
and adjusting the output min and max ranges accordingly.</p><dl><dt>input_min, input_max</dt><dd>are scalar floats that specify the range for the float
interpretation of the <code>input</code> data. For example, if input_min is -1.0f and
input_max is 1.0f, and we are dealing with quint16 quantized data, then a 0
value in the 16-bit data should be interpreted as -1.0f, and a 65535 means 1.0f.</dd></dl><p>This operator tries to squeeze as much precision as possible into an output with
a lower bit depth by calculating the actual min and max values found in the
data. For example, maybe that quint16 input has no values lower than 16,384 and
none higher than 49,152. That means only half the range is actually needed, all
the float interpretations are between -0.5f and 0.5f, so if we want to compress
the data into a quint8 output, we can use that range rather than the theoretical
-1.0f to 1.0f that is suggested by the input min and max.</p><p>In practice, this is most useful for taking output from operations like
QuantizedMatMul that can produce higher bit-depth outputs than their inputs and
may have large potential output ranges, but in practice have a distribution of
input values that only uses a small fraction of the possible range. By feeding
that output into this operator, we can reduce it from 32 bits down to 8 with
minimal loss of accuracy.</p></div></div><divclass="top"><pclass="src"><aname="v:quantizedMatMul"class="def">quantizedMatMul</a><ahref="src/TensorFlow-GenOps-Core.html#quantizedMatMul"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t1, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` t1, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t2, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` t2, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> toutput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` toutput)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t1</td><tdclass="doc"><p><strong>a</strong>: Must be a two-dimensional tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t2</td><tdclass="doc"><p><strong>b</strong>: Must be a two-dimensional tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>min_a</strong>: The float value that the lowest quantized <code>a</code> value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>max_a</strong>: The float value that the highest quantized <code>a</code> value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>min_b</strong>: The float value that the lowest quantized <code>b</code> value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>max_b</strong>: The float value that the highest quantized <code>b</code> value represents.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> toutput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>out</strong>, <strong>min_out</strong>, <strong>max_out</strong>)</p><ul><li><strong>out</strong></li><li><strong>min_out</strong>: The float value that the lowest quantized output value represents.</li><li><strong>max_out</strong>: The float value that the highest quantized output value represents.</li></ul>
<code>a</code> (after being transposed if <code>transpose_a</code> is non-zero) must match the
outer dimension of <code>b</code> (after being transposed if <code>transposed_b</code> is
non-zero).</p></div></div><divclass="top"><pclass="src"><aname="v:cumprod"class="def">cumprod</a><ahref="src/TensorFlow-GenOps-Core.html#cumprod"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>axis</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>out</strong></p></td></tr></table></div><divclass="doc"><p>Compute the cumulative product of the tensor <code>x</code> along <code>axis</code>.</p><p>By default, this op performs an inclusive cumprod, which means that the first
element of the input is identical to the first element of the output:
```prettyprint
tf.cumprod([a, b, c]) ==> [a, a * b, a * b * c]
```</p><p>By setting the <code>exclusive</code> kwarg to <code><ahref="../base-4.8.2.0/Data-Bool.html#v:True">True</a></code>, an exclusive cumprod is
performed instead:
```prettyprint
tf.cumprod([a, b, c], exclusive=True) ==> [0, a, a * b]
```</p><p>By setting the <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:reverse">reverse</a></code> kwarg to <code><ahref="../base-4.8.2.0/Data-Bool.html#v:True">True</a></code>, the cumprod is performed in the
opposite direction:
```prettyprint
tf.cumprod([a, b, c], reverse=True) ==> [a * b * c, b * c, c]
```
This is more efficient than using separate `tf.reverse` ops.</p><p>The <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:reverse">reverse</a></code> and <code>exclusive</code> kwargs can also be combined:
```</p></div></div><divclass="top"><pclass="src"><aname="v:cumsum"class="def">cumsum</a><ahref="src/TensorFlow-GenOps-Core.html#cumsum"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>axis</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>out</strong></p></td></tr></table></div><divclass="doc"><p>Compute the cumulative sum of the tensor <code>x</code> along <code>axis</code>.</p><p>By default, this op performs an inclusive cumsum, which means that the first
element of the input is identical to the first element of the output:
```prettyprint
tf.cumsum([a, b, c]) ==> [a, a + b, a + b + c]
```</p><p>By setting the <code>exclusive</code> kwarg to <code><ahref="../base-4.8.2.0/Data-Bool.html#v:True">True</a></code>, an exclusive cumsum is
performed instead:
```prettyprint
tf.cumsum([a, b, c], exclusive=True) ==> [0, a, a + b]
```</p><p>By setting the <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:reverse">reverse</a></code> kwarg to <code><ahref="../base-4.8.2.0/Data-Bool.html#v:True">True</a></code>, the cumsum is performed in the
opposite direction:
```prettyprint
tf.cumsum([a, b, c], reverse=True) ==> [a + b + c, b + c, c]
```
This is more efficient than using separate `tf.reverse` ops.</p><p>The <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:reverse">reverse</a></code> and <code>exclusive</code> kwargs can also be combined:
```</p></div></div><divclass="top"><pclass="src"><aname="v:cross"class="def">cross</a><ahref="src/TensorFlow-GenOps-Core.html#cross"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>a</strong>: A tensor containing 3-element vectors.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>b</strong>: Another tensor, of same type and shape as <code>a</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>product</strong>: Pairwise cross product of the vectors in <code>a</code> and <code>b</code>.</p></td></tr></table></div><divclass="doc"><p>Compute the pairwise cross product.</p><p><code>a</code> and <code>b</code> must be the same shape; they can either be simple 3-element vectors,
or any shape where the innermost dimension is 3. In the latter case, each pair
of corresponding 3-element vectors is cross-multiplied independently.</p></div></div><divclass="top"><pclass="src"><aname="v:iFFT3D"class="def">iFFT3D</a><ahref="src/TensorFlow-GenOps-Core.html#iFFT3D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>input</strong>: A complex64 tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>output</strong>: A complex64 tensor of the same shape as <code>input</code>. The inner-most 3
dimensions of <code>input</code> are replaced with their inverse 3D Fourier Transform.</p><p><code>compatibility(numpy)
Equivalent to np.fft3
</code>end_compatibility</p></td></tr></table></div><divclass="doc"><p>Compute the inverse 3-dimensional discrete Fourier Transform over the inner-most</p><p>3 dimensions of <code>input</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:fFT3D"class="def">fFT3D</a><ahref="src/TensorFlow-GenOps-Core.html#fFT3D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>input</strong>: A complex64 tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>output</strong>: A complex64 tensor of the same shape as <code>input</code>. The inner-most 3
dimensions of <code>input</code> are replaced with their 3D Fourier Transform.</p><p><code>compatibility(numpy)
Equivalent to np.fft3
</code>end_compatibility</p></td></tr></table></div><divclass="doc"><p>Compute the 3-dimensional discrete Fourier Transform over the inner-most 3</p><p>dimensions of <code>input</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:maxPoolGradWithArgmax"class="def">maxPoolGradWithArgmax</a><ahref="src/TensorFlow-GenOps-Core.html#maxPoolGradWithArgmax"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> targmax, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` targmax, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: The original input.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>grad</strong>: 4-D with shape `[batch, height, width, channels]`. Gradients w.r.t. the
output of <code>max_pool</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 targmax</td><tdclass="doc"><p><strong>argmax</strong>: The indices of the maximum values chosen for each output of <code>max_pool</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Gradients w.r.t. the input of <code>max_pool</code>.</p></td></tr></table></div><divclass="doc"><p>Computes gradients of the maxpooling function.</p></div></div><divclass="top"><pclass="src"><aname="v:fFT2D"class="def">fFT2D</a><ahref="src/TensorFlow-GenOps-Core.html#fFT2D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>input</strong>: A complex64 tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>output</strong>: A complex64 tensor of the same shape as <code>input</code>. The inner-most 2
dimensions of <code>input</code> are replaced with their 2D Fourier Transform.</p><p><code>compatibility(numpy)
Equivalent to np.fft2
</code>end_compatibility</p></td></tr></table></div><divclass="doc"><p>Compute the 2-dimensional discrete Fourier Transform over the inner-most</p><p>2 dimensions of <code>input</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:iFFT"class="def">iFFT</a><ahref="src/TensorFlow-GenOps-Core.html#iFFT"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>input</strong>: A complex64 tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>output</strong>: A complex64 tensor of the same shape as <code>input</code>. The inner-most
dimension of <code>input</code> is replaced with its inverse 1D Fourier Transform.</p></td></tr></table></div><divclass="doc"><p>Compute the inverse 1-dimensional discrete Fourier Transform over the inner-most</p><p>dimension of <code>input</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:fFT"class="def">fFT</a><ahref="src/TensorFlow-GenOps-Core.html#fFT"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>input</strong>: A complex64 tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>output</strong>: A complex64 tensor of the same shape as <code>input</code>. The inner-most
dimension of <code>input</code> is replaced with its 1D Fourier Transform.</p></td></tr></table></div><divclass="doc"><p>Compute the 1-dimensional discrete Fourier Transform over the inner-most</p><p>dimension of <code>input</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:conj"class="def">conj</a><ahref="src/TensorFlow-GenOps-Core.html#conj"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Returns the complex conjugate of a complex number.</p><p>Given a tensor <code>input</code> of complex numbers, this operation returns a tensor of
complex numbers that are the complex conjugate of each element in <code>input</code>. The
complex numbers in <code>input</code> must be of the form \(a + bj\), where *a* is the
real part and *b* is the imaginary part.</p><p>The complex conjugate returned by this operation is of the form \(a - bj\).</p><p>For example:</p><p>```
# tensor <code>input</code> is [-2.25 + 4.75j, 3.25 + 5.75j]
```</p></div></div><divclass="top"><pclass="src"><aname="v:real"class="def">real</a><ahref="src/TensorFlow-GenOps-Core.html#real"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tout, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` tout)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> tout</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Returns the real part of a complex number.</p><p>Given a tensor <code>input</code> of complex numbers, this operation returns a tensor of
type <code>float</code> that is the real part of each element in <code>input</code>. All elements in
<code>input</code> must be complex numbers of the form \(a + bj\), where *a* is the real
part returned by this operation and *b* is the imaginary part.</p><p>For example:</p><p>```
# tensor <code>input</code> is [-2.25 + 4.75j, 3.25 + 5.75j]
tf.real(input) ==> [-2.25, 3.25]
```</p></div></div><divclass="top"><pclass="src"><aname="v:complex"class="def">complex</a><ahref="src/TensorFlow-GenOps-Core.html#complex"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tout, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` tout)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>real</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>imag</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> tout</td><tdclass="doc"><p><strong>out</strong></p></td></tr></table></div><divclass="doc"><p>Converts two real numbers to a complex number.</p><p>Given a tensor <code><ahref="TensorFlow-GenOps-Core.html#v:real">real</a></code> representing the real part of a complex number, and a
tensor <code><ahref="TensorFlow-GenOps-Core.html#v:imag">imag</a></code> representing the imaginary part of a complex number, this
operation returns complex numbers elementwise of the form \(a + bj\), where
*a* represents the <code><ahref="TensorFlow-GenOps-Core.html#v:real">real</a></code> part and *b* represents the <code><ahref="TensorFlow-GenOps-Core.html#v:imag">imag</a></code> part.</p><p>The input tensors <code><ahref="TensorFlow-GenOps-Core.html#v:real">real</a></code> and <code><ahref="TensorFlow-GenOps-Core.html#v:imag">imag</a></code> must have the same shape.</p><p>For example:</p><p>```
# tensor <code><ahref="TensorFlow-GenOps-Core.html#v:real">real</a></code> is [2.25, 3.25]
# tensor <code><ahref="TensorFlow-GenOps-Core.html#v:imag">imag</a></code> is [4.75, 5.75]
```</p></div></div><divclass="top"><pclass="src"><aname="v:range"class="def">range</a><ahref="src/TensorFlow-GenOps-Core.html#range"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 tidx</td><tdclass="doc"><p><strong>start</strong>: 0-D (scalar). First entry in the sequence.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>limit</strong>: 0-D (scalar). Upper limit of sequence, exclusive.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 tidx</td><tdclass="doc"><p><strong>delta</strong>: 0-D (scalar). Optional. Default is 1. Number that increments <code>start</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> tidx</td><tdclass="doc"><p><strong>output</strong>: 1-D.</p></td></tr></table></div><divclass="doc"><p>Creates a sequence of numbers.</p><p>This operation creates a sequence of numbers that begins at <code>start</code> and
extends by increments of <code>delta</code> up to but not including <code>limit</code>.</p><p>For example:</p><p>```
```</p></div></div><divclass="top"><pclass="src"><aname="v:any"class="def">any</a><ahref="src/TensorFlow-GenOps-Core.html#any"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>input</strong>: The tensor to reduce.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>reduction_indices</strong>: The dimensions to reduce.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>output</strong>: The reduced tensor.</p></td></tr></table></div><divclass="doc"><p>Computes the "logical or" of elements across dimensions of a tensor.</p><p>Reduces <code>input</code> along the dimensions given in <code>reduction_indices</code>. Unless
<code>keep_dims</code> is true, the rank of the tensor is reduced by 1 for each entry in
<code>reduction_indices</code>. If <code>keep_dims</code> is true, the reduced dimensions are
retained with length 1.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseSegmentMean"class="def">sparseSegmentMean</a><ahref="src/TensorFlow-GenOps-Core.html#sparseSegmentMean"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>data</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>indices</strong>: A 1-D tensor. Has same rank as <code>segment_ids</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>segment_ids</strong>: A 1-D tensor. Values should be sorted and can be repeated.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Has same shape as data, except for dimension 0 which
has size <code>k</code>, the number of segments.</p></td></tr></table></div><divclass="doc"><p>Computes the mean along sparse segments of a tensor.</p><p>Read <ahref="../../api_docs/python/math_ops.md#segmentation">the section on
Segmentation</a> for an explanation
of segments.</p><p>Like <code>SegmentMean</code>, but <code>segment_ids</code> can have rank less than `data`'s first
dimension, selecting a subset of dimension 0, specified by <code>indices</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseSegmentSum"class="def">sparseSegmentSum</a><ahref="src/TensorFlow-GenOps-Core.html#sparseSegmentSum"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>data</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>indices</strong>: A 1-D tensor. Has same rank as <code>segment_ids</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>segment_ids</strong>: A 1-D tensor. Values should be sorted and can be repeated.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Has same shape as data, except for dimension 0 which
has size <code>k</code>, the number of segments.</p></td></tr></table></div><divclass="doc"><p>Computes the sum along sparse segments of a tensor.</p><p>Read <ahref="../../api_docs/python/math_ops.md#segmentation">the section on
Segmentation</a> for an explanation
of segments.</p><p>Like <code>SegmentSum</code>, but <code>segment_ids</code> can have rank less than `data`'s first
dimension, selecting a subset of dimension 0, specified by <code>indices</code>.</p><p>For example:</p><p>```prettyprint
c = tf.constant([[1,2,3,4], [-1,-2,-3,-4], [5,6,7,8]])</p><p># Select two rows, one segment.
```</p></div></div><divclass="top"><pclass="src"><aname="v:unsortedSegmentSum"class="def">unsortedSegmentSum</a><ahref="src/TensorFlow-GenOps-Core.html#unsortedSegmentSum"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>data</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>segment_ids</strong>: A tensor whose shape is a prefix of `data.shape`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>num_segments</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Has same shape as data, except for the first `segment_ids.rank`
dimensions, which are replaced with a single dimension which has size
<code>num_segments</code>.</p></td></tr></table></div><divclass="doc"><p>Computes the sum along segments of a tensor.</p><p>Read <ahref="../../api_docs/python/math_ops.md#segmentation">the section on
Segmentation</a> for an explanation
of segments.</p><p>Computes a tensor such that
`(output[i] = sum_{j...} data[j...]` where the sum is over tuples `j...` such
that `segment_ids[j...] == i`. Unlike <code>SegmentSum</code>, <code>segment_ids</code>
need not be sorted and need not cover all values in the full
range of valid values.</p><p>If the sum is empty for a given segment ID <code>i</code>, `output[i] = 0`.</p><p><code>num_segments</code> should equal the number of distinct segment IDs.</p><p><ahref="div">style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;"</a>
<ahref="/div">/div</a></p></div></div><divclass="top"><pclass="src"><aname="v:segmentMin"class="def">segmentMin</a><ahref="src/TensorFlow-GenOps-Core.html#segmentMin"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>data</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>segment_ids</strong>: A 1-D tensor whose rank is equal to the rank of `data`'s
first dimension. Values should be sorted and can be repeated.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Has same shape as data, except for dimension 0 which
has size <code>k</code>, the number of segments.</p></td></tr></table></div><divclass="doc"><p>Computes the minimum along segments of a tensor.</p><p>Read <ahref="../../api_docs/python/math_ops.md#segmentation">the section on
Segmentation</a> for an explanation
of segments.</p><p>Computes a tensor such that
\(output_i = min_j(data_j)\) where <code><ahref="../base-4.8.2.0/Data-Ord.html#v:min">min</a></code> is over <code>j</code> such
that `segment_ids[j] == i`.</p><p><ahref="div">style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;"</a>
<ahref="/div">/div</a></p></div></div><divclass="top"><pclass="src"><aname="v:segmentProd"class="def">segmentProd</a><ahref="src/TensorFlow-GenOps-Core.html#segmentProd"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>data</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>segment_ids</strong>: A 1-D tensor whose rank is equal to the rank of `data`'s
first dimension. Values should be sorted and can be repeated.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Has same shape as data, except for dimension 0 which
has size <code>k</code>, the number of segments.</p></td></tr></table></div><divclass="doc"><p>Computes the product along segments of a tensor.</p><p>Read <ahref="../../api_docs/python/math_ops.md#segmentation">the section on
Segmentation</a> for an explanation
of segments.</p><p>Computes a tensor such that
\(output_i = prod_j data_j\) where the product is over <code>j</code> such
that `segment_ids[j] == i`.</p><p><ahref="div">style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;"</a>
<ahref="/div">/div</a></p></div></div><divclass="top"><pclass="src"><aname="v:segmentMean"class="def">segmentMean</a><ahref="src/TensorFlow-GenOps-Core.html#segmentMean"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>data</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>segment_ids</strong>: A 1-D tensor whose rank is equal to the rank of `data`'s
first dimension. Values should be sorted and can be repeated.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Has same shape as data, except for dimension 0 which
has size <code>k</code>, the number of segments.</p></td></tr></table></div><divclass="doc"><p>Computes the mean along segments of a tensor.</p><p>Read <ahref="../../api_docs/python/math_ops.md#segmentation">the section on
Segmentation</a> for an explanation
of segments.</p><p>Computes a tensor such that
\(output_i = frac{sum_j data_j}{N}\) where <code><ahref="TensorFlow-GenOps-Core.html#v:mean">mean</a></code> is
over <code>j</code> such that `segment_ids[j] == i` and <code>N</code> is the total number of
<ahref="/div">/div</a></p></div></div><divclass="top"><pclass="src"><aname="v:segmentSum"class="def">segmentSum</a><ahref="src/TensorFlow-GenOps-Core.html#segmentSum"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>data</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>segment_ids</strong>: A 1-D tensor whose rank is equal to the rank of `data`'s
first dimension. Values should be sorted and can be repeated.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Has same shape as data, except for dimension 0 which
has size <code>k</code>, the number of segments.</p></td></tr></table></div><divclass="doc"><p>Computes the sum along segments of a tensor.</p><p>Read <ahref="../../api_docs/python/math_ops.md#segmentation">the section on Segmentation</a>
for an explanation of segments.</p><p>Computes a tensor such that
\(output_i = sum_j data_j\) where sum is over <code>j</code> such
that `segment_ids[j] == i`.</p><p><ahref="div">style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;"</a>
of the input Tensor to reduce across. For vectors, use dimension = 0.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Returns the index with the smallest value across dimensions of a tensor.</p></div></div><divclass="top"><pclass="src"><aname="v:max"class="def">max</a><ahref="src/TensorFlow-GenOps-Core.html#max"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: The tensor to reduce.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>reduction_indices</strong>: The dimensions to reduce.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The reduced tensor.</p></td></tr></table></div><divclass="doc"><p>Computes the maximum of elements across dimensions of a tensor.</p><p>Reduces <code>input</code> along the dimensions given in <code>reduction_indices</code>. Unless
<code>keep_dims</code> is true, the rank of the tensor is reduced by 1 for each entry in
<code>reduction_indices</code>. If <code>keep_dims</code> is true, the reduced dimensions are
retained with length 1.</p></div></div><divclass="top"><pclass="src"><aname="v:min"class="def">min</a><ahref="src/TensorFlow-GenOps-Core.html#min"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: The tensor to reduce.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>reduction_indices</strong>: The dimensions to reduce.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The reduced tensor.</p></td></tr></table></div><divclass="doc"><p>Computes the minimum of elements across dimensions of a tensor.</p><p>Reduces <code>input</code> along the dimensions given in <code>reduction_indices</code>. Unless
<code>keep_dims</code> is true, the rank of the tensor is reduced by 1 for each entry in
<code>reduction_indices</code>. If <code>keep_dims</code> is true, the reduced dimensions are
retained with length 1.</p></div></div><divclass="top"><pclass="src"><aname="v:prod"class="def">prod</a><ahref="src/TensorFlow-GenOps-Core.html#prod"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: The tensor to reduce.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>reduction_indices</strong>: The dimensions to reduce.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The reduced tensor.</p></td></tr></table></div><divclass="doc"><p>Computes the product of elements across dimensions of a tensor.</p><p>Reduces <code>input</code> along the dimensions given in <code>reduction_indices</code>. Unless
<code>keep_dims</code> is true, the rank of the tensor is reduced by 1 for each entry in
<code>reduction_indices</code>. If <code>keep_dims</code> is true, the reduced dimensions are
retained with length 1.</p></div></div><divclass="top"><pclass="src"><aname="v:sum"class="def">sum</a><ahref="src/TensorFlow-GenOps-Core.html#sum"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: The tensor to reduce.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>reduction_indices</strong>: The dimensions to reduce.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The reduced tensor.</p></td></tr></table></div><divclass="doc"><p>Computes the sum of elements across dimensions of a tensor.</p><p>Reduces <code>input</code> along the dimensions given in <code>reduction_indices</code>. Unless
<code>keep_dims</code> is true, the rank of the tensor is reduced by 1 for each entry in
<code>reduction_indices</code>. If <code>keep_dims</code> is true, the reduced dimensions are
retained with length 1.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseMatMul"class="def">sparseMatMul</a><ahref="src/TensorFlow-GenOps-Core.html#sparseMatMul"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> ta, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` ta, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tb, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` tb)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 ta</td><tdclass="doc"><p><strong>a</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tb</td><tdclass="doc"><p><strong>b</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>product</strong></p></td></tr></table></div><divclass="doc"><p>Multiply matrix "a" by matrix "b".</p><p>The inputs must be two-dimensional matrices and the inner dimension of "a" must
match the outer dimension of "b". This op is optimized for the case where at
least one of "a" or "b" is sparse. The breakeven for using this versus a dense
matrix multiply on one platform was 30% zero values in the sparse matrix.</p></div></div><divclass="top"><pclass="src"><aname="v:matMul"class="def">matMul</a><ahref="src/TensorFlow-GenOps-Core.html#matMul"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>a</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>b</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>product</strong></p></td></tr></table></div><divclass="doc"><p>Multiply the matrix "a" by the matrix "b".</p><p>The inputs must be two-dimensional matrices and the inner dimension of
"a" (after being transposed if transpose_a is true) must match the
outer dimension of "b" (after being transposed if transposed_b is
true).</p><ul><li>Note*: The default kernel implementation for MatMul on GPUs uses
cublas.</li></ul></div></div><divclass="top"><pclass="src"><aname="v:logicalAnd"class="def">logicalAnd</a><ahref="src/TensorFlow-GenOps-Core.html#logicalAnd"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns the truth value of x AND y element-wise.</p><ul><li>NOTE*: <code>LogicalAnd</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:equal"class="def">equal</a><ahref="src/TensorFlow-GenOps-Core.html#equal"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a>, <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns the truth value of (x == y) element-wise.</p><ul><li>NOTE*: <code>Equal</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:greaterEqual"class="def">greaterEqual</a><ahref="src/TensorFlow-GenOps-Core.html#greaterEqual"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns the truth value of (x >= y) element-wise.</p><ul><li>NOTE*: <code>GreaterEqual</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:lessEqual"class="def">lessEqual</a><ahref="src/TensorFlow-GenOps-Core.html#lessEqual"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns the truth value of (x <= y) element-wise.</p><ul><li>NOTE*: <code>LessEqual</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:less"class="def">less</a><ahref="src/TensorFlow-GenOps-Core.html#less"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns the truth value of (x < y) element-wise.</p><ul><li>NOTE*: <code>Less</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:polygamma"class="def">polygamma</a><ahref="src/TensorFlow-GenOps-Core.html#polygamma"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>a</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Compute the polygamma function \(psi^{(n)}(x)\).</p><p>The polygamma function is defined as:</p><p>```
psi^{(n)}(x) = frac{d^n}{dx^n} psi(x)
```
where \(psi(x)\) is the digamma function.</p></div></div><divclass="top"><pclass="src"><aname="v:igamma"class="def">igamma</a><ahref="src/TensorFlow-GenOps-Core.html#igamma"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>a</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Compute the lower regularized incomplete Gamma function `Q(a, x)`.</p><p>The lower regularized incomplete Gamma function is defined as:</p><p>```
P(a, x) = gamma(a, x) / Gamma(a) = 1 - Q(a, x)
```
where
```
gamma(a, x) = int_{0}^{x} t^{a-1} exp(-t) dt
```
is the lower incomplete Gamma function.</p><p>Note, above `Q(a, x)` (<code>Igammac</code>) is the upper regularized complete
Gamma function.</p></div></div><divclass="top"><pclass="src"><aname="v:igammac"class="def">igammac</a><ahref="src/TensorFlow-GenOps-Core.html#igammac"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>a</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Compute the upper regularized incomplete Gamma function `Q(a, x)`.</p><p>The upper regularized incomplete Gamma function is defined as:</p><p>```
Q(a, x) = Gamma(a, x) / Gamma(a) = 1 - P(a, x)
```
where
```
Gamma(a, x) = int_{x}^{infty} t^{a-1} exp(-t) dt
```
is the upper incomplete Gama function.</p><p>Note, above `P(a, x)` (<code>Igamma</code>) is the lower regularized complete
Gamma function.</p></div></div><divclass="top"><pclass="src"><aname="v:mod"class="def">mod</a><ahref="src/TensorFlow-GenOps-Core.html#mod"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns element-wise remainder of division.</p><ul><li>NOTE*: <code>Mod</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:minimum"class="def">minimum</a><ahref="src/TensorFlow-GenOps-Core.html#minimum"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns the min of x and y (i.e. x < y ? x : y) element-wise.</p><ul><li>NOTE*: <code>Minimum</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:maximum"class="def">maximum</a><ahref="src/TensorFlow-GenOps-Core.html#maximum"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns the max of x and y (i.e. x > y ? x : y) element-wise.</p><ul><li>NOTE*: <code>Maximum</code> supports broadcasting. More about broadcasting
width = width_pad - crop_left - crop_right</p><p>The attr <code>block_size</code> must be greater than one. It indicates the block size.</p><p>Some examples:</p><ol><li>For the following input of shape `[4, 1, 1, 1]` and block_size of 2:</li></ol><p>```prettyprint
[[[[1]]], [[[2]]], [[[3]]], [[[4]]]]
```</p><p>The output tensor has shape `[1, 2, 2, 1]` and value:</p><p>```prettyprint
x = [[[[1], [2]], [[3], [4]]]]
```</p><ol><li>For the following input of shape `[4, 1, 1, 3]` and block_size of 2:</li></ol><p>```prettyprint
```</p><p>The output tensor has shape `[2, 2, 4, 1]` and value:</p><p>```prettyprint
x = [[[[1], [3]], [[5], [7]]],
[[[2], [4]], [[10], [12]]],
[[[5], [7]], [[13], [15]]],
[[[6], [8]], [[14], [16]]]]
```</p></td></tr></table></div><divclass="doc"><p>BatchToSpace for 4-D tensors of type T.</p><p>This is a legacy version of the more general BatchToSpaceND.</p><p>Rearranges (permutes) data from batch into blocks of spatial data, followed by
cropping. This is the reverse transformation of SpaceToBatch. More specifically,
this op outputs a copy of the input tensor where values from the <code>batch</code>
dimension are moved in spatial blocks to the <code>height</code> and <code>width</code> dimensions,
followed by cropping along the <code>height</code> and <code>width</code> dimensions.</p></div></div><divclass="top"><pclass="src"><aname="v:mul"class="def">mul</a><ahref="src/TensorFlow-GenOps-Core.html#mul"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns x * y element-wise.</p><ul><li>NOTE*: <code>Mul</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:rint"class="def">rint</a><ahref="src/TensorFlow-GenOps-Core.html#rint"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Returns element-wise integer closest to x.</p><p>If the result is midway between two representable values,
```</p></div></div><divclass="top"><pclass="src"><aname="v:ceil"class="def">ceil</a><ahref="src/TensorFlow-GenOps-Core.html#ceil"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Returns element-wise smallest integer in not less than x.</p></div></div><divclass="top"><pclass="src"><aname="v:floor"class="def">floor</a><ahref="src/TensorFlow-GenOps-Core.html#floor"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Returns element-wise largest integer not greater than x.</p></div></div><divclass="top"><pclass="src"><aname="v:maxPool3D"class="def">maxPool3D</a><ahref="src/TensorFlow-GenOps-Core.html#maxPool3D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Shape `[batch, depth, rows, cols, channels]` tensor to pool over.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The max pooled output tensor.</p></td></tr></table></div><divclass="doc"><p>Performs 3D max pooling on the input.</p></div></div><divclass="top"><pclass="src"><aname="v:isInf"class="def">isInf</a
Equivalent to np.isinf
</code>end_compatibility</p></div></div><divclass="top"><pclass="src"><aname="v:depthwiseConv2dNativeBackpropInput"class="def">depthwiseConv2dNativeBackpropInput</a><ahref="src/TensorFlow-GenOps-Core.html#depthwiseConv2dNativeBackpropInput"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>input_sizes</strong>: An integer vector representing the shape of <code>input</code>,
where <code>input</code> is a 4-D `[batch, height, width, channels]` tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>filter</strong>: 4-D with shape
Gradients w.r.t. the output of the convolution.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 4-D with shape `[batch, in_height, in_width, in_channels]`. Gradient
w.r.t. the input of the convolution.</p></td></tr></table></div><divclass="doc"><p>Computes the gradients of depthwise convolution with respect to the input.</p></div></div><divclass="top"><pclass="src"><aname="v:isNan"class="def">isNan</a><ahref="src/TensorFlow-GenOps-Core.html#isNan"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Returns which elements of x are NaN.</p><p><code>compatibility(numpy)
Equivalent to np.isnan
</code>end_compatibility</p></div></div><divclass="top"><pclass="src"><aname="v:log1p"class="def">log1p</a><ahref="src/TensorFlow-GenOps-Core.html#log1p"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes natural logarithm of (1 + x) element-wise.</p><p>I.e., \(y = log_e (1 + x)\).</p></div></div><divclass="top"><pclass="src"><aname="v:asin"class="def">asin</a><ahref="src/TensorFlow-GenOps-Core.html#asin"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes asin of x element-wise.</p></div></div><divclass="top"><pclass="src"><aname="v:topKV2"class="def">topKV2</a><ahref="src/TensorFlow-GenOps-Core.html#topKV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: 1-D or higher with last dimension at least <code>k</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorfl
row for matrices).</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>)</td><tdclass="doc"><p>(<strong>values</strong>, <strong>indices</strong>)</p><ul><li><strong>values</strong>: The <code>k</code> largest elements along each last dimensional slice.</li><li><strong>indices</strong>: The indices of <code>values</code> within the last dimension of <code>input</code>.</li></ul></td></tr></table></div><divclass="doc"><p>Finds values and indices of the <code>k</code> largest elements for the last dimension.</p><p>If the input is a vector (rank-1), finds the <code>k</code> largest entries in the vector
and outputs their values and indices as vectors. Thus `values[j]` is the
<code>j</code>-th largest entry in <code>input</code>, and its index is `indices[j]`.</p><p>For matrices (resp. higher rank input), computes the top <code>k</code> entries in each
row (resp. vector along the last dimension). Thus,</p><p>values.shape = indices.shape = input.shape[:-1] + [k]</p><p>If two elements are equal, the lower-index element appears first.</p><p>This is the same as <code>TopK</code>, but takes <code>k</code> as in input rather than an attr.</p></div></div><divclass="top"><pclass="src"><aname="v:cos"class="def">cos</a><ahref="src/TensorFlow-GenOps-Core.html#cos"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes cos of x element-wise.</p></div></div><divclass="top"><pclass="src"><aname="v:sin"class="def">sin</a><ahref="src/TensorFlow-GenOps-Core.html#sin"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes sin of x element-wise.</p></div></div><divclass="top"><pclass="src"><aname="v:randomUniformInt"class="def">randomUniformInt</a><ahref="src/TensorFlow-GenOps-Core.html#randomUniformInt"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tout, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tout, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>shape</strong>: The shape of the output tensor.</p></td></tr><tr><tdclass="src">-><ahref
The lower bound <code>minval</code> is included in the range, while the upper bound
<code>maxval</code> is excluded.</p><p>The random integers are slightly biased unless `maxval - minval` is an exact
power of two. The bias is small for values of `maxval - minval` significantly
smaller than the range of the output (either `2^32` or `2^64`).</p></div></div><divclass="top"><pclass="src"><aname="v:erfc"class="def">erfc</a><ahref="src/TensorFlow-GenOps-Core.html#erfc"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes the complementary error function of <code>x</code> element-wise.</p></div></div><divclass="top"><pclass="src"><aname="v:digamma"class="def">digamma</a><ahref="src/TensorFlow-GenOps-Core.html#digamma"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes Psi, the derivative of Lgamma (the log of the absolute value of</p><p>`Gamma(x)`), element-wise.</p></div></div><divclass="top"><pclass="src"><aname="v:fusedResizeAndPadConv2D"class="def">fusedResizeAndPadConv2D</a><ahref="src/TensorFlow-GenOps-Core.html#fusedResizeAndPadConv2D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: 4-D with shape `[batch, in_height, in_width, in_channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>size</strong>: A 1-D int32 Tensor of 2 elements: `new_height, new_width`. The
new size for the images.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>paddings</strong>: A two-column matrix specifying the padding sizes. The number of
rows must be the same as the rank of <code>input</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>filter</strong>: 4-D with shape
`[filter_height, filter_width, in_channels, out_channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Performs a resize and padding as a preprocess during a convolution.</p><p>It's often possible to do spatial transformations more efficiently as part of
the packing stage of a convolution, so this op allows for an optimized
implementation where these stages are fused together. This prevents the need to
write out the intermediate results as whole tensors, reducing memory pressure,
and we can get some latency gains by merging the transformation calculations.
The data_format attribute for Conv2D isn't supported by this op, and defaults to
<code>NHWC</code> order.
Internally this op uses a single per-graph scratch buffer, which means that it
will block if multiple versions are being run in parallel. This is because this
operator is primarily an optimization to minimize memory usage.</p></div></div><divclass="top"><pclass="src"><aname="v:sub"class="def">sub</a><ahref="src/TensorFlow-GenOps-Core.html#sub"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns x - y element-wise.</p><ul><li>NOTE*: <code>Sub</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:sign"class="def">sign</a><ahref="src/TensorFlow-GenOps-Core.html#sign"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Returns an element-wise indication of the sign of a number.</p><p>`y = sign(x) = -1` if `x <ahref="0`;">0 if `x == 0`; 1 if `x</a> 0`.</p><p>For complex numbers, `y = sign(x) = x / |x|` if `x != 0`, otherwise `y = 0`.</p></div></div><divclass="top"><pclass="src"><aname="v:lgamma"class="def">lgamma</a><ahref="src/TensorFlow-GenOps-Core.html#lgamma"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes the log of the absolute value of `Gamma(x)` element-wise.</p></div></div><divclass="top"><pclass="src"><aname="v:log"class="def">log</a><ahref="src/TensorFlow-GenOps-Core.html#log"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><t
<code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code> tensor has shape `[filter_height, filter_width, depth]`, i.e., each
input channel is processed independently of the others with its own structuring
function. The <code>output</code> tensor has shape
`[batch, out_height, out_width, depth]`. The spatial dimensions of the output
tensor depend on the <code>padding</code> algorithm. We currently only support the default
<ahref="NHWC.html">NHWC</a><code>data_format</code>.</p><p>In detail, the grayscale morphological 2-D dilation is the max-sum correlation
(for consistency with <code>conv2d</code>, we use unmirrored filters):</p><p>output[b, y, x, c] =
max_{dy, dx} input[b,
strides[1] * y + rates[1] * dy,
strides[2] * x + rates[2] * dx,
c] +
filter[dy, dx, c]</p><p>Max-pooling is a special case when the filter has size equal to the pooling
kernel size and contains all zeros.</p><p>Note on duality: The dilation of <code>input</code> by the <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code> is equal to the
negation of the erosion of `-input` by the reflected <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code>.</p></div></div><divclass="top"><pclass="src"><aname="v:rsqrtGrad"class="def">rsqrtGrad</a><ahref="src/TensorFlow-GenOps-Core.html#rsqrtGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Computes the gradient for the rsqrt of <code>x</code> wrt its input.</p><p>Specifically, `grad = dy * -0.5 * y^3`, where `y = rsqrt(x)`, and <code>dy</code>
is the corresponding input gradient.</p></div></div><divclass="top"><pclass="src"><aname="v:rsqrt"class="def">rsqrt</a><ahref="src/TensorFlow-GenOps-Core.html#rsqrt"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes reciprocal of square root of x element-wise.</p><p>I.e., \(y = 1 / sqrt{x}\).</p></div></div><divclass="top"><pclass="src"><aname="v:quantizedMaxPool"class="def">quantizedMaxPool</a><ahref="src/TensorFlow-GenOps-Core.html#quantizedMaxPool"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: The 4D (batch x rows x cols x depth) Tensor to MaxReduce over.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>min_input</strong>: The float value that the lowest quantized input value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>max_input</strong>: The float value that the highest quantized input value represents.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>output</strong>, <strong>min_output</strong>, <strong>max_output</strong>)</p><ul><li><strong>output</strong></li><li><strong>min_output</strong>: The float value that the lowest quantized output value represents.</li><li><strong>max_output</strong>: The float value that the highest quantized output value represents.</li></ul></td></tr></table></div><divclass="doc"><p>Produces the max pool of the input tensor for quantized types.</p></div></div><divclass="top"><pclass="src"><aname="
work string and output (work, work).</p></div></div><divclass="top"><pclass="src"><aname="v:square"class="def">square</a><ahref="src/TensorFlow-GenOps-Core.html#square"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes square of x element-wise.</p><p>I.e., \(y = x * x = x^2\).</p></div></div><divclass="top"><pclass="src"><aname="v:quantizedReshape"class="def">quantizedReshape</a><ahref="src/TensorFlow-GenOps-Core.html#quantizedReshape"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tshape, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tshape)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>tensor</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tshape</td><tdclass="doc"><p><strong>shape</strong>: Defines the shape of the output tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>input_min</strong>: The minimum value of the input.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>input_max</strong>: The maximum value of the input.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>output</strong>, <strong>output_min</strong>, <strong>output_max</strong>)</p><ul><li><strong>output</strong></li><li><strong>output_min</strong>: This value is copied from input_min.</li><li><strong>output_max</strong>: This value is copied from input_max.</li></ul></td></tr></table></div><divclass="doc"><p>Reshapes a quantized tenso
is the corresponding input gradient.</p></div></div><divclass="top"><pclass="src"><aname="v:invGrad"class="def">invGrad</a><ahref="src/TensorFlow-GenOps-Core.html#invGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Computes the gradient for the inverse of <code>x</code> wrt its input.</p><p>Specifically, `grad = -dy * y*y`, where `y = 1/x`, and <code>dy</code>
is the corresponding input gradient.</p></div></div><divclass="top"><pclass="src"><aname="v:inv"class="def">inv</a><ahref="src/TensorFlow-GenOps-Core.html#inv"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes the reciprocal of x element-wise.</p><p>I.e., \(y = 1 / x\).</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArrayConcatV2"class="def">tensorArrayConcatV2</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArrayConcatV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a TensorArray.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong>: A float scalar that enforces proper chaining of operations.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p>(<strong>value</strong>, <strong>lengths</strong>)</p><ul><li><strong>value</strong>: All of the elements in the TensorArray, concatenated along the first
axis.</li><li><strong>lengths</strong>: A vector of the row sizes of the original T elements in the
value output. In the example above, this would be the values:
`(n1, n2, ..., n(T-1))`.</li></ul></td></tr></table></div><divclass="doc"><p>Concat the elements from the TensorArray into value <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>.</p><p>Takes <code>T</code> elements of shapes</p><p>```
(n0 x d0 x d1 x ...), (n1 x d0 x d1 x ...), ..., (n(T-1) x d0 x d1 x ...)
```</p><p>and concatenates them into a Tensor of shape:</p><p>```(n0 + n1 + ... + n(T-1) x d0 x d1 x ...)```</p><p>All elements must have the same shape (excepting the first dimension).</p></div></div><divclass="top"><pclass="src"><aname="v:complexAbs"class="def">complexAbs</a><ahref="src/TensorFlow-GenOps-Core.html#complexAbs"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tout, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` tout)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> tout</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes the complex absolute value of a tensor.</p><p>Given a tensor <code>x</code> of complex numbers, this operation returns a tensor of type
<code>float</code> or <code>double</code> that is the absolute value of each element in <code>x</code>. All
elements in <code>x</code> must be complex numbers of the form \(a + bj\). The absolute
value is computed as \( sqrt{a^2 + b^2}\).</p><p>For example:</p><p>```
# tensor <code>x</code> is [[-2.25 + 4.75j], [-3.25 + 5.75j]]
tf.complex_abs(x) ==> [5.25594902, 6.60492229]
```</p></div></div><divclass="top"><pclass="src"><aname="v:_HostCast"class="def">_HostCast</a><ahref="src/TensorFlow-GenOps-Core.html#_HostCast"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> srcT, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dstT)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 srcT</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dstT</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Cast x of type SrcT to y of DstT.</p><p>_HostCast requires its input and produces its output in host memory.</p></div></div><divclass="top"><pclass="src"><aname="v:resizeNearestNeighbor"class="def">resizeNearestNeighbor</a><ahref="src/TensorFlow-GenOps-Core.html#resizeNearestNeighbor"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>images</strong>: 4-D with shape `[batch, height, width, channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>size</strong>: = A 1-D int32 Tensor of 2 elements: `new_height, new_width`. The
new size for the images.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>resized_images</strong>: 4-D with shape
`sum_per_d(gradients * (inputs > max))`.</li></ul></td></tr></table></div><divclass="doc"><p>Compute gradients for a FakeQuantWithMinMaxVarsPerChannel operation.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseSegmentSqrtNGrad"class="def">sparseSegmentSqrtNGrad</a><ahref="src/TensorFlow-GenOps-Core.html#sparseSegmentSqrtNGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>grad</strong>: gradient propagated to the SparseSegmentSqrtN op.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>indices</strong>: indices passed to the corresponding SparseSegmentSqrtN op.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>segment_ids</strong>: segment_ids passed to the corresponding SparseSegmentSqrtN op.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>output_dim0</strong>: dimension 0 of "data" passed to SparseSegmentSqrtN op.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Computes gradients for SparseSegmentSqrtN.</p><p>Returns tensor "output" with same shape as grad, except for dimension 0 whose
value is output_dim0.</p></div></div><divclass="top"><pclass="src"><aname="v:fakeQuantWithMinMaxVarsPerChannel"class="def">fakeQuantWithMinMaxVarsPerChannel</a><ahref="src/TensorFlow-GenOps-Core.html#fakeQuantWithMinMaxVarsPerChannel"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>inputs</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>min</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>max</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>outputs</strong></p></td></tr></table></div><divclass="doc"><p>Fake-quantize the <code>inputs</code> tensor of type float and one of the shapes: `[d]`,</p><p>`[b, d]` `[b, h, w, d]` via per-channel floats <code><ahref="../base-4.8.2.0/Data-Ord.html#v:min">min</a></code> and <code><ahref="../base-4.8.2.0/Data-Ord.html#v:max">max</a></code> of shape `[d]`
to <code>outputs</code> tensor of same shape as <code>inputs</code>.</p><dl><dt>min; max</dt><dd>is the clamping range for the <code>inputs</code> data in the corresponding
depth channel. Op divides this range into 255 steps (total of 256 values), then
replaces each <code>inputs</code> value with the closest of the quantized step values.</dd></dl><p>This operation has a gradient and thus allows for training <code><ahref="../base-4.8.2.0/Data-Ord.html#v:min">min</a></code> and <code><ahref="../base-4.8.2.0/Data-Ord.html#v:max">max</a></code> values.</p></div></div><divclass="top"><pclass="src"><aname="v:scalarSummary"class="def">scalarSummary</a><ahref="src/TensorFlow-GenOps-Core.html#scalarSummary"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>tags</strong>: Tags for the summary.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>values</strong>: Same shape as `tags. Values for the summary.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>summary</strong>: Scalar. Serialized <code>Summary</code> protocol buffer.</p></td></tr></table></div><divclass="doc"><p>Outputs a <code>Summary</code> protocol buffer with scalar values.</p><p>The input <code>tags</code> and <code>values</code> must have the same shape. The generated summary
has a summary value for each tag-value pair in <code>tags</code> and <code>values</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:neg"class="def">neg</a><ahref="src/TensorFlow-GenOps-Core.html#neg"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes numerical negative value element-wise.</p><p>I.e., \(y = -x\).</p></div></div><divclass="top"><pclass="src"><aname="v:fakeQuantWithMinMaxArgsGradient"class="def">fakeQuantWithMinMaxArgsGradient</a><ahref="src/TensorFlow-GenOps-Core.html#fakeQuantWithMinMaxArgsGradient"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>gradients</strong>: Backpropagated gradients above the FakeQuantWithMinMaxArgs operation.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>inputs</strong>: Values passed as inputs to the FakeQuantWithMinMaxArgs operation.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>backprops</strong>: Backpropagated gradients below the FakeQuantWithMinMaxArgs operation:
`gradients * (inputs >= min && inputs <= max)`.</p></td></tr></table></div><divclass="doc"><p>Compute gradients for a FakeQuantWithMinMaxArgs operation.</p></div></div><divclass="top"><pclass="src"><aname="v:debugNanCount"class="def">debugNanCount</a><ahref="src/TensorFlow-GenOps-Core.html#debugNanCount"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Input tensor, non-Reference type.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>output</strong>: An integer output tensor that is the number of NaNs in the input.</p></td></tr></table></div><divclass="doc"><p>Debug NaN Value Counter Op</p><p>Counts number of NaNs in the input tensor, for debugging.</p></div></div><divclass="top"><pclass="src"><aname="v:debugIdentity"class="def">debugIdentity</a><ahref="src/TensorFlow-GenOps-Core.html#debugIdentity"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Input tensor, non-Reference type.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Output tensor that equals the input tensor.</p></td></tr></table></div><divclass="doc"><p>Debug Identity Op.</p><p>Provides an identity mapping of the non-Ref type input tensor for debugging.</p></div></div><divclass="top"><pclass="src"><aname="v:bitcast"class="def">bitcast</a><ahref="src/TensorFlow-GenOps-Core.html#bitcast"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> type', <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Dat
data as <code>input</code> with datatype `type`.</p><p>If the input datatype <code>T</code> is larger than the output datatype `type` then the
shape changes from [...] to [..., sizeof(<code>T</code>)/sizeof(`type`)].</p><p>If <code>T</code> is smaller than `type`, the operator requires that the rightmost
dimension be equal to sizeof(`type`)/sizeof(<code>T</code>). The shape then goes from
[..., sizeof(`type`)/sizeof(<code>T</code>)] to [...].</p><ul><li>NOTE*: Bitcast is implemented as a low-level cast, so machines with different
endian orderings will give different results.</li></ul></div></div><divclass="top"><pclass="src"><aname="v:sigmoid"class="def">sigmoid</a><ahref="src/TensorFlow-GenOps-Core.html#sigmoid"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes sigmoid of <code>x</code> element-wise.</p><p>Specifically, `y = 1 / (1 + exp(-x))`.</p></div></div><divclass="top"><pclass="src"><aname="v:copy"class="def">copy</a><ahref="src/TensorFlow-GenOps-Core.html#copy"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Input tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Output tensor, deep-copied from input.</p></td></tr></table></div><divclass="doc"><p>Copy Op.</p><p>Performs CPU-to-CPU or GPU-to-GPU deep-copying of tensor, depending on the
device on which the tensor is allocated.</p><p>Unlike the CopyHost Op, this op does not have HostMemory constraint on its
input or output.</p></div></div><divclass="top"><pclass="src"><aname="v:fixedUnigramCandidateSampler"class="def">fixedUnigramCandidateSampler</a><ahref="src/TensorFlow-GenOps-Core.html#fixedUnigramCandidateSampler"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_sampled</strong>: Number of candidates to randomly sample per batch.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_true</strong>: Number of true labels per context.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>range_max</strong>: The sampler will sample integers from the interval [0, range_max).</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>unique</strong>: If unique is true, we sample with rejection, so that all sampled
candidates in a batch are unique. This requires some approximation to
estimate the post-rejection sampling probabilities.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>true_classes</strong>: A batch_size * num_true matrix, in which each row contains the
IDs of the num_true target_classes in the corresponding original label.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>sampled_candidates</strong>, <strong>true_expected_count</strong>, <strong>sampled_expected_count</strong>)</p><ul><li><strong>sampled_candidates</strong>: A vector of length num_sampled, in which each element is
the ID of a sampled candidate.</li><li><strong>true_expected_count</strong>: A batch_size * num_true matrix, representing
the number of times each candidate is expected to occur in a batch
of sampled candidates. If unique=true, then this is a probability.</li><li><strong>sampled_expected_count</strong>: A vector of length num_sampled, for each sampled
candidate representing the number of times the candidate is expected
to occur in a batch of sampled candidates. If unique=true, then this is a
probability.</li></ul></td></tr></table></div><divclass="doc"><p>Generates labels for candidate sampling with a learned unigram distribution.</p><p>A unigram sampler could use a fixed unigram distribution read from a
file or passed in as an in-memory array instead of building up the distribution
from data on the fly. There is also an option to skew the distribution by
applying a distortion power to the weights.</p><p>The vocabulary file should be in CSV-like format, with the last field
being the weight associated with the word.</p><p>For each batch, this op picks a single set of sampled candidate labels.</p><p>The advantages of sampling candidates per-batch are simplicity and the
possibility of efficient dense matrix multiplication. The disadvantage is that
the sampled candidates must be chosen independently of the context and of the
true labels.</p></div></div><divclass="top"><pclass="src"><aname="v:listDiff"class="def">listDiff</a><ahref="src/TensorFlow-GenOps-Core.html#listDiff"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> out_idx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` out_idx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong>: 1-D. Values to keep.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong>: 1-D. Values to remove.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_idx)</td><tdclass="doc"><p>(<strong>out</strong>, <strong>idx</strong>)</p><ul><li><strong>out</strong>: 1-D. Values present in <code>x</code> but not in <code>y</code>.</li><li><strong>idx</strong>: 1-D. Positions of <code>x</code> values preserved in <code>out</code>.</li></ul></td></tr></table></div><divclass="doc"><p>Computes the difference between two lists of numbers or strings.</p><p>Given a list <code>x</code> and a list <code>y</code>, this operation returns a list <code>out</code> that
represents all values that are in <code>x</code> but not in <code>y</code>. The returned list <code>out</code>
is sorted in the same order that the numbers appear in <code>x</code> (duplicates are
preserved). This operation also returns a list <code>idx</code> that represents the
position of each <code>out</code> element in <code>x</code>. In other words:</p><p>`out[i] = x[idx[i]] for i in [0, 1, ..., len(out) - 1]`</p><p>For example, given this input:</p><p>```prettyprint
x = [1, 2, 3, 4, 5, 6]
y = [1, 3, 5]
```</p><p>This operation would return:</p><p>```prettyprint
ksize_cols * depth]` containing image patches with size
`ksize_rows x ksize_cols x depth` vectorized in the "depth" dimension.</p></td></tr></table></div><divclass="doc"><p>Extract <code>patches</code> from <code>images</code> and put them in the "depth" output dimension.</p></div></div><divclass="top"><pclass="src"><aname="v:spaceToDepth"class="def">spaceToDepth</a><ahref="src/TensorFlow-GenOps-Core.html#spaceToDepth"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>block_size</strong>: The size of the spatial block.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>SpaceToDepth for tensors of type T.</p><p>Rearranges blocks of spatial data, into depth. More specifically,
this op outputs a copy of the input tensor where values from the <code>height</code>
and <code>width</code> dimensions are moved to the <code>depth</code> dimension.
The attr <code>block_size</code> indicates the input block size and how the data is moved.</p><ul><li>Non-overlapping blocks of size `block_size x block size` are rearranged
into depth at each location.</li><li>The depth of the output tensor is `input_depth * block_size * block_size`.</li><li>The input tensor's height and width must be divisible by block_size.</li></ul><p>That is, assuming the input is in the shape:
`[batch, height, width, depth]`,
the shape of the output will be:
`[batch, height<em>block_size, width</em>block_size, depth*block_size*block_size]`</p><p>This operation requires that the input tensor be of rank 4, and that
<code>block_size</code> be >=1 and a divisor of both the input <code>height</code> and <code>width</code>.</p><p>This operation is useful for resizing the activations between convolutions
(but keeping all data), e.g. instead of pooling. It is also useful for training
purely convolutional models.</p><p>For example, given this input of shape `[1, 2, 2, 1]`, and block_size of 2:</p><p>```prettyprint
x = [[[[1], [2]],
[[3], [4]]]]
```</p><p>This operation will output a tensor of shape `[1, 1, 1, 4]`:</p><p>```prettyprint
[[[[1, 2, 3, 4]]]]
```</p><p>Here, the input has a batch of 1 and each batch element has shape `[2, 2, 1]`,
the corresponding output will have a single element (i.e. width and height are
both 1) and will have a depth of 4 channels (1 * block_size * block_size).
The output element shape is `[1, 1, 4]`.</p><p>For an input tensor with larger depth, here of shape `[1, 2, 2, 3]`, e.g.</p><p>```prettyprint
x = [[[[1, 2, 3], [4, 5, 6]],
[[7, 8, 9], [10, 11, 12]]]]
```</p><p>This operation, for block_size of 2, will return the following tensor of shape
`[1, 1, 1, 12]`</p><p>```prettyprint
[[[[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12]]]]
```</p><p>Similarly, for the following input of shape `[1 4 4 1]`, and a block size of 2:</p><p>```prettyprint
x = [[[[1], [2], [5], [6]],
[[3], [4], [7], [8]],
[[9], [10], [13], [14]],
[[11], [12], [15], [16]]]]
```</p><p>the operator will return the following tensor of shape `[1 2 2 4]`:</p><p>```prettyprint
x = [[[[1, 2, 3, 4],
[5, 6, 7, 8]],
[[9, 10, 11, 12],
[13, 14, 15, 16]]]]
```</p></div></div><divclass="top"><pclass="src"><aname="v:cropAndResizeGradBoxes"class="def">cropAndResizeGradBoxes</a><ahref="src/TensorFlow-GenOps-Core.html#cropAndResizeGradBoxes"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>grads</strong>: A 4-D tensor of shape `[num_boxes, crop_height, crop_width, depth]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>image</strong>: A 4-D tensor of shape `[batch, image_height, image_width, depth]`.
Both <code>image_height</code> and <code>image_width</code> need to be positive.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>boxes</strong>: A 2-D tensor of shape `[num_boxes, 4]`. The <code>i</code>-th row of the tensor
<code>extrapolation_value</code> to extrapolate the input image values.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>box_ind</strong>: A 1-D tensor of shape `[num_boxes]` with int32 values in `[0, batch)`.
The value of `box_ind[i]` specifies the image that the <code>i</code>-th box refers to.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>output</strong>: A 2-D tensor of shape `[num_boxes, 4]`.</p></td></tr></table></div><divclass="doc"><p>Computes the gradient of the crop_and_resize op wrt the input boxes tensor.</p></div></div><divclass="top"><pclass="src"><aname="v:batchToSpaceND"class="def">batchToSpaceND</a><ahref="src/TensorFlow-GenOps-Core.html#batchToSpaceND"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tblock_shape, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tblock_shape, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tcrops, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tcrops)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: N-D with shape `input_shape = [batch] + spatial_shape + remaining_shape`,
where spatial_shape has M dimensions.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tblock_shape</td><tdclass="doc"><p><strong>block_shape</strong>: 1-D with shape `[M]`, all values must be >= 1.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 tcrops</td><tdclass="doc"><p><strong>crops</strong>: 2-D with shape `[M, 2]`, all values must be >= 0.
`crops[i] = [crop_start, crop_end]` specifies the amount to crop from input
dimension `i + 1`, which corresponds to spatial dimension <code>i</code>. It is
required that
`crop_start[i] + crop_end[i] <= block_shape[i] * input_shape[i + 1]`.</p><p>This operation is equivalent to the following steps:</p><ol><li>Reshape <code>input</code> to <code>reshaped</code> of shape:
[block_shape[0], ..., block_shape[M-1],
batch / prod(block_shape),
input_shape[1], ..., input_shape[N-1]]</li><li>Permute dimensions of <code>reshaped</code> to produce <code>permuted</code> of shape
input_shape[M], block_shape[M-1],</p><p>input_shape[M+1], ..., input_shape[N-1]]</p><ol><li>Reshape <code>permuted</code> to produce <code>reshaped_permuted</code> of shape
```</p><p>The output tensor has shape `[2, 2, 4, 1]` and value:</p><p>```prettyprint
x = [[[[1], [2], [3], [4]],
[[5], [6], [7], [8]]],
[[[9], [10], [11], [12]],
[[13], [14], [15], [16]]]]
```</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>BatchToSpace for N-D tensors of type T.</p><p>This operation reshapes the "batch" dimension 0 into `M + 1` dimensions of shape
`block_shape + [batch]`, interleaves these blocks back into the grid defined by
the spatial dimensions `[1, ..., M]`, to obtain a result with the same rank as
the input. The spatial dimensions of this intermediate result are then
optionally cropped according to <code>crops</code> to produce the output. This is the
reverse of SpaceToBatch. See below for a precise description.</p></div></div><divclass="top"><pclass="src"><aname="v:spaceToBatch"class="def">spaceToBatch</a><ahref="src/TensorFlow-GenOps-Core.html#spaceToBatch"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tpaddings, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tpaddings)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>block_size</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: 4-D with shape `[batch, height, width, depth]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tpaddings</td><tdclass="doc"><p><strong>paddings</strong>: 2-D tensor of non-negative integers with shape `[2, 2]`. It specifies
the padding of the input with zeros across the spatial dimensions as follows:</p><p>paddings = [[pad_top, pad_bottom], [pad_left, pad_right]]</p><p>The effective spatial dimensions of the zero-padded input tensor will be:</p><p>height_pad = pad_top + height + pad_bottom
width_pad = pad_left + width + pad_right</p><p>The attr <code>block_size</code> must be greater than one. It indicates the block size.</p><ul><li>Non-overlapping blocks of size `block_size x block size` in the height and
width dimensions are rearranged into the batch dimension at each location.</li><li>The batch of the output tensor is `batch * block_size * block_size`.</li><li>Both height_pad and width_pad must be divisible by block_size.</li></ul><p>The shape of the output will be:</p><p>[batch*block_size*block_size, height_pad<em>block_size, width_pad</em>block_size,
depth]</p><p>Some examples:</p><ol><li>For the following input of shape `[1, 2, 2, 1]` and block_size of 2:</li></ol><p>```prettyprint
x = [[[[1], [2]], [[3], [4]]]]
```</p><p>The output tensor has shape `[4, 1, 1, 1]` and value:</p><p>```prettyprint
[[[[1]]], [[[2]]], [[[3]]], [[[4]]]]
```</p><ol><li>For the following input of shape `[1, 2, 2, 3]` and block_size of 2:</li></ol><p>```prettyprint
x = [[[[1, 2, 3], [4, 5, 6]],
[[7, 8, 9], [10, 11, 12]]]]
```</p><p>The output tensor has shape `[4, 1, 1, 3]` and value:</p><p>```prettyprint
```</p><p>Among others, this operation is useful for reducing atrous convolution into
regular convolution.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>SpaceToBatch for 4-D tensors of type T.</p><p>This is a legacy version of the more general SpaceToBatchND.</p><p>Zero-pads and then rearranges (permutes) blocks of spatial data into batch.
More specifically, this op outputs a copy of the input tensor where values from
the <code>height</code> and <code>width</code> dimensions are moved to the <code>batch</code> dimension. After
the zero-padding, both <code>height</code> and <code>width</code> of the input must be divisible by the
block size.</p></div></div><divclass="top"><pclass="src"><aname="v:adjustHue"class="def">adjustHue</a><ahref="src/TensorFlow-GenOps-Core.html#adjustHue"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>images</strong>: Images to adjust. At least 3-D.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>delta</strong>: A float delta to add to the hue.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>output</strong>: The hue-adjusted image or images.</p></td></tr></table></div><divclass="doc"><p>Adjust the hue of one or more images.</p><p><code>images</code> is a tensor of at least 3 dimensions. The last dimension is
interpretted as channels, and must be three.</p><p>The input image is considered in the RGB colorspace. Conceptually, the RGB
colors are first mapped into HSV. A delta is then applied all the hue values,
and then remapped back to RGB colorspace.</p></div></div><divclass="top"><pclass="src"><aname="v:spaceToBatchND"class="def">spaceToBatchND</a><ahref="src/TensorFlow-GenOps-Core.html#spaceToBatchND"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tblock_shape, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tblock_shape, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tpaddings, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tpaddings)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: N-D with shape `input_shape = [batch] + spatial_shape + remaining_shape`,
where spatial_shape has <code>M</code> dimensions.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tblock_shape</td><tdclass="doc"><p><strong>block_shape</strong>: 1-D with shape `[M]`, all values must be >= 1.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 tpaddings</td><tdclass="doc"><p><strong>paddings</strong>: 2-D with shape `[M, 2]`, all values must be >= 0.
`paddings[i] = [pad_start, pad_end]` specifies the padding for input dimension
`i + 1`, which corresponds to spatial dimension <code>i</code>. It is required that
`block_shape[i]` divides `input_shape[i + 1] + pad_start + pad_end`.</p><p>This operation is equivalent to the following steps:</p><ol><li>Zero-pad the start and end of dimensions `[1, ..., M]` of the
input according to <code>paddings</code> to produce <code>padded</code> of shape <code>padded_shape</code>.</li><li>Reshape <code>padded</code> to <code>reshaped_padded</code> of shape:</li></ol><dl><dt>batch</dt><dd>+</dd><dt>padded_shape[1</dt><dd>/ block_shape[0],
block_shape[0],
...,
padded_shape[M] / block_shape[M-1],
block_shape[M-1]] +
remaining_shape</dd></dl><ol><li>Permute dimensions of <code>reshaped_padded</code> to produce
<code>permuted_reshaped_padded</code> of shape:</li></ol><p>block_shape +
[batch] +
[padded_shape[1] / block_shape[0],
...,
padded_shape[M] / block_shape[M-1]] +
remaining_shape</p><ol><li>Reshape <code>permuted_reshaped_padded</code> to flatten <code>block_shape</code> into the batch
dimension, producing an output tensor of shape:</li></ol><dl><dt>batch * prod(block_shape)</dt><dd>+</dd><dt>padded_shape[1</dt><dd>/ block_shape[0],
...,
padded_shape[M] / block_shape[M-1]] +
remaining_shape</dd></dl><p>Some examples:</p><ol><li>For the following input of shape `[1, 2, 2, 1]`, `block_shape = [2, 2]`, and
```</p><p>The output tensor has shape `[8, 1, 3, 1]` and value:</p><p>```prettyprint
x = [[[[0], [1], [3]]], [[[0], [9], [11]]],
[[[0], [2], [4]]], [[[0], [10], [12]]],
[[[0], [5], [7]]], [[[0], [13], [15]]],
[[[0], [6], [8]]], [[[0], [14], [16]]]]
```</p><p>Among others, this operation is useful for reducing atrous convolution into
regular convolution.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>SpaceToBatch for N-D tensors of type T.</p><p>This operation divides "spatial" dimensions `[1, ..., M]` of the input into a
grid of blocks of shape <code>block_shape</code>, and interleaves these blocks with the
"batch" dimension (0) such that in the output, the spatial dimensions
`[1, ..., M]` correspond to the position within the grid, and the batch
dimension combines both the position within a spatial block and the original
batch position. Prior to division into blocks, the spatial dimensions of the
input are optionally zero padded according to <code>paddings</code>. See below for a
precise description.</p></div></div><divclass="top"><pclass="src"><aname="v:diagPart"class="def">diagPart</a><ahref="src/TensorFlow-GenOps-Core.html#diagPart"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Rank k tensor where k is 2, 4, or 6.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>diagonal</strong>: The extracted diagonal.</p></td></tr></table></div><divclass="doc"><p>Returns the diagonal part of the tensor.</p><p>This operation returns a tensor with the <code>diagonal</code> part
of the <code>input</code>. The <code>diagonal</code> part is computed as follows:</p><p>Assume <code>input</code> has dimensions `[D1,..., Dk, D1,..., Dk]`, then the output is a
tensor of rank <code>k</code> with dimensions `[D1,..., Dk]` where:</p><p>`diagonal[i1,..., ik] = input[i1, ..., ik, i1,..., ik]`.</p><p>For example:</p><p>```prettyprint
```</p></div></div><divclass="top"><pclass="src"><aname="v:placeholderV2"class="def">placeholderV2</a><ahref="src/TensorFlow-GenOps-Core.html#placeholderV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:Shape">Shape</a></td><tdclass="doc"><p><strong>shape</strong>: The shape of the tensor. The shape can be any partially-specified
shape. To be unconstrained, pass in a shape with unknown rank.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype</td><tdclass="doc"><p><strong>output</strong>: A placeholder tensor that must be replaced using the feed mechanism.</p></td></tr></table></div><divclass="doc"><p>A placeholder op for a value that will be fed into the computation.</p><p>N.B. This operation will fail with an error if it is executed. It is
intended as a way to represent a value that will always be fed, and to
provide attrs that enable the fed value to be checked at runtime.</p></div></div><divclass="top"><pclass="src"><aname="v:acos"class="def">acos</a><ahref="src/TensorFlow-GenOps-Core.html#acos"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes acos of x element-wise.</p></div></div><divclass="top"><pclass="src"><aname="v:placeholder"class="def">placeholder</a><ahref="src/TensorFlow-GenOps-Core.html#placeholder"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype</td><tdclass="doc"><p><strong>output</strong>: A placeholder tensor that must be replaced using the feed mechanism.</p></td></tr></table></div><divclass="doc"><p>A placeholder op for a value that will be fed into the computation.</p><p>N.B. This operation will fail with an error if it is executed. It is
intended as a way to represent a value that will always be fed, and to
provide attrs that enable the fed value to be checked at runtime.</p></div></div><divclass="top"><pclass="src"><aname="v:controlTrigger"class="def">controlTrigger</a> :: <ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a><ahref="src/TensorFlow-GenOps-Core.html#controlTrigger"class="link">Source</a></p><divclass="doc"><p>Does nothing. Serves as a control trigger for scheduling.</p><p>Only useful as a placeholder for control edges.</p></div></div><divclass="top"><pclass="src"><aname="v:atan"class="def">atan</a><ahref="src/TensorFlow-GenOps-Core.html#atan"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes atan of x element-wise.</p></div></div><divclass="top"><pclass="src"><aname="v:mirrorPad"class="def">mirrorPad</a><ahref="src/TensorFlow-GenOps-Core.html#mirrorPad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tpaddings, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tpaddings)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: The input tensor to be padded.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tpaddings</td><tdclass="doc"><p><strong>paddings</strong>: A two-column matrix specifying the padding sizes. The number of
rows must be the same as the rank of <code>input</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The padded tensor.</p></td></tr></table></div><divclass="doc"><p>Pads a tensor with mirrored values.</p><p>This operation pads a <code>input</code> with mirrored values according to the <code>paddings</code>
you specify. <code>paddings</code> is an integer tensor with shape `[n, 2]`, where n is
the rank of <code>input</code>. For each dimension D of <code>input</code>, `paddings[D, 0]` indicates
how many values to add before the contents of <code>input</code> in that dimension, and
`paddings[D, 1]` indicates how many values to add after the contents of <code>input</code>
in that dimension. Both `paddings[D, 0]` and `paddings[D, 1]` must be no greater
than `input.dim_size(D)` (or `input.dim_size(D) - 1`) if <code>copy_border</code> is true
(if false, respectively).</p><p>The padded size of each dimension D of the output is:</p><p>`paddings(D, 0) + input.dim_size(D) + paddings(D, 1)`</p><p>For example:</p><p>```prettyprint
# <code>t</code> is [[1, 2, 3], [4, 5, 6]].
# <code>paddings</code> is [[1, 1]], [2, 2]].
# <code>mode</code> is SYMMETRIC.
# rank of <code>t</code> is 2.
pad(t, paddings) ==> [[2, 1, 1, 2, 3, 3, 2]
[2, 1, 1, 2, 3, 3, 2]
[5, 4, 4, 5, 6, 6, 5]
[5, 4, 4, 5, 6, 6, 5]]
```</p></div></div><divclass="top"><pclass="src"><aname="v:where-39-"class="def">where'</a><ahref="src/TensorFlow-GenOps-Core.html#where%27"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>index</strong></p></td></tr></table></div><divclass="doc"><p>Returns locations of true values in a boolean tensor.</p><p>This operation returns the coordinates of true elements in <code>input</code>. The
```</p></div></div><divclass="top"><pclass="src"><aname="v:avgPool3DGrad"class="def">avgPool3DGrad</a><ahref="src/TensorFlow-GenOps-Core.html#avgPool3DGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>orig_input_shape</strong>: The original input dimensions.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>grad</strong>: Output backprop of shape `[batch, depth, rows, cols, channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The backprop for input.</p></td></tr></table></div><divclass="doc"><p>Computes gradients of average pooling function.</p></div></div><divclass="top"><pclass="src"><aname="v:readerReset"class="def">readerReset</a><ahref="src/TensorFlow-GenOps-Core.html#readerReset"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>reader_handle</strong>: Handle to a Reader.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Restore a Reader to its initial clean state.</p></div></div><divclass="top"><pclass="src"><aname="v:tileGrad"class="def">tileGrad</a><ahref="src/TensorFlow-GenOps-Core.html#tileGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>multiples</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Returns the gradient of <code>Tile</code>.</p><p>Since <code>Tile</code> takes an input and repeats the input <c
each repeated tile of <code>input</code> into <code>output</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:expandDims"class="def">expandDims</a><ahref="src/TensorFlow-GenOps-Core.html#expandDims"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tdim, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tdim)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tdim</td><tdclass="doc"><p><strong>dim</strong>: 0-D (scalar). Specifies the dimension index at which to
expand the shape of <code>input</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Contains the same data as <code>input</code>, but its shape has an additional
dimension of size 1 added.</p></td></tr></table></div><divclass="doc"><p>Inserts a dimension of 1 into a tensor's shape.</p><p>Given a tensor <code>input</code>, this operation inserts a dimension of 1 at the
dimension index <code>dim</code> of <code>input</code>'s shape. The dimension index <code>dim</code> starts at
zero; if you specify a negative number for <code>dim</code> it is counted backward from
the end.</p><p>This operation is useful if you want to add a batch dimension to a single
element. For example, if you have a single image of shape `[height, width,
channels]`, you can make it a batch of 1 image with `expand_dims(image, 0)`,
which will make the shape `[1, height, width, channels]`.</p><p>Other examples:</p><p>```prettyprint
# <code>t</code> is a tensor of shape [2]
shape(expand_dims(t, 0)) ==> [1, 2]
shape(expand_dims(t, 1)) ==> [2, 1]
shape(expand_dims(t, -1)) ==> [2, 1]</p><p># <code>t2</code> is a tensor of shape [2, 3, 5]
shape(expand_dims(t2, 0)) ==> [1, 2, 3, 5]
shape(expand_dims(t2, 2)) ==> [2, 3, 1, 5]
shape(expand_dims(t2, 3)) ==> [2, 3, 5, 1]
```</p><p>This operation requires that:</p><p>`-1-input.dims() <= dim <= input.dims()`</p><p>This operation is related to `squeeze()`, which removes dimensions of
size 1.</p></div></div><divclass="top"><pclass="src"><aname="v:tensorSummary"class="def">tensorSummary</a><ahref="src/TensorFlow-GenOps-Core.html#tensorSummary"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>tensor</strong>: A tensor to serialize.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>summary</strong></p></td></tr></table></div><divclass="doc"><p>Outputs a <code>Summary</code> protocol buffer with a tensor.</p></div></div><divclass="top"><pclass="src"><aname="v:tile"class="def">tile</a><ahref="src/TensorFlow-GenOps-Core.html#tile"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tmultiples, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tmultiples)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: 1-D or higher.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tmultiples</td><tdclass="doc"><p><strong>multiples</strong>: 1-D. Length must be the same as the number of dimensions in <code>input</code></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Constructs a tensor by tiling a given tensor.</p><p>This operation creates a new tensor by replicating <code>input</code><code>multiples</code> times.
The output tensor's i'th dimension has `input.dims(i) * multiples[i]` elements,
and the values of <code>input</code> are replicated `multiples[i]` times along the <code>i</code>th
dimension. For example, tiling `[a b c d]` by `[2]` produces
`[a b c d a b c d]`.</p></div></div><divclass="top"><pclass="src"><aname="v:stridedSlice"class="def">stridedSlice</a><ahref="src/TensorFlow-GenOps-Core.html#stridedSlice"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> index, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` index)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 index</td><tdclass="doc"><p><strong>begin</strong>: `begin[k]` specifies the offset into the <code>k</code>th range specification.
The exact dimension this corresponds to will be determined by context.
Out-of-bounds values will be silently clamped. If the <code>k</code>th bit of
<code>begin_mask</code> then `begin[k]` is ignored and the full range of the
appropriate dimension is used instead. Negative values causes indexing
to start from the highest element e.g. If `foo==[1,2,3]` then `foo[-1]==3`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 index</td><tdclass="doc"><p><strong>end</strong>: `end[i]` is like <code>begin</code> with the exception that <code>end_mask</code> is
used to determine full ranges.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 index</td><tdclass="doc"><p><strong>strides</strong>: `strides[i]` specifies the increment in the <code>i</code>th specification
after extracting a given element. Negative indices will reverse
the original order. Out or range values are
clamped to `[0,dim[i]) if slice[i]>0` or `[-1,dim[i]-1] if slice[i] < 0`</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Return a strided slice from <code>input</code>.</p><p>Note, most python users will want to use the Python <code><ahref="Tensor.html#v:__getitem__">__getitem__</a></code>
or <code><ahref="Variable.html#v:__getitem__">__getitem__</a></code> rather than this op directly.</p><p>The goal of this op is to produce a new tensor with a subset of
the elements from the <code>n</code> dimensional <code>input</code> tensor. The subset is chosen using
a sequence of <code>m</code> sparse range specifications encoded into the arguments
of this function. Note, in some cases
<code>m</code> could be equal to <code>n</code>, but this need not be the case. Each
range specification entry can be one of the following:</p><ul><li>An ellipsis (...). Ellipses are used to imply zero or more
dimensions of full-dimension selection and are produced using
<code>ellipsis_mask</code>. For example, `foo[...]` is the identity slice.</li><li>A new axis. This is used to insert a new shape=1 dimension and is
produced using <code>new_axis_mask</code>. For example, `foo[:, ...]` where
<code>foo</code> is shape `(3, 4)` produces a `(1, 3, 4)` tensor.</li><li>A range `begin:end:stride`. This is used to specify how much to choose from
a given dimension. <code>stride</code> can be any integer but 0. <code>begin</code> is an integer
which represents the index of the first value to select while <code>end</code> represents
the index of the last value to select. The number of values selected in each
dimension is `end - begin` if `stride > 0` and `begin - end` if `stride < 0`.
<code>begin</code> and <code>end</code> can be negative where `-1` is the last element, `-2` is
the second to last. <code>begin_mask</code> controls whether to replace the explicitly
given <code>begin</code> with an implicit effective value of `0` if `stride > 0` and
`-1` if `stride < 0`. <code>end_mask</code> is analogous but produces the number
required to create the largest open interval. For example, given a shape
`(3,)` tensor `foo[:]`, the effective <code>begin</code> and <code>end</code> are `0` and `3`. Do
not assume this is equivalent to `foo[0:-1]` which has an effective <code>begin</code>
and <code>end</code> of `0` and `2`. Another example is `foo[-2::-1]` which reverses the
first dimension of a tensor while dropping the last two (in the original
order elements). For example `foo = [1,2,3,4]; foo[-2::-1]` is `[4,3]`.</li><li>A single index. This is used to keep only elements that have a given
index. For example (`foo[2, :]` on a shape `(5,6)` tensor produces a
shape `(6,)` tensor. This is encoded in <code>begin</code> and <code>end</code> and
<code>shrink_axis_mask</code>.</li></ul><p>Each conceptual range specification is encoded in the op's argument. This
encoding is best understand by considering a non-trivial example. In
particular,
`foo[1, 2:4, None, ..., :-3:-1, :]` will be encoded as</p><p>```prettyprint
begin = [1, 2, x, x, 0, x] # x denotes don't care (usually 0)
end = [2, 4, x, x, -3, x]
strides = [1, 1, x, x, -1, 1]
begin_mask = 1<<4 | 1 << 5 = 48
end_mask = 1<<5 = 32
ellipsis_mask = 1<<3 = 8
new_axis_mask = 1<<2 4
shrink_axis_mask = 1<<0
```</p><p>In this case if `foo.shape` is (5, 5, 5, 5, 5, 5) the final shape of
the slice becomes (2, 1, 5, 5, 2, 5).
Let us walk step by step through each argument specification.</p><ol><li>The first argument in the example slice is turned into `begin = 1` and
`end = begin + 1 = 2`. To disambiguate from the original spec `2:4` we
also set the appropriate bit in <code>shrink_axis_mask</code>.</li><li>`2:4` is contributes 2, 4, 1 to begin, end, and stride. All masks have
zero bits contributed.</li><li>None is a synonym for `tf.newaxis`. This means insert a dimension of size 1
dimension in the final shape. Dummy values are contributed to begin,
end and stride, while the new_axis_mask bit is set.</li><li><code>...</code> grab the full ranges from as many dimensions as needed to
fully specify a slice for every dimension of the input shape.</li><li>`:-3:-1` shows the use of negative indices. A negative index <code>i</code> associated
with a dimension that has shape <code>s</code> is converted to a positive index
`s + i`. So `-1` becomes `s-1` (i.e. the last element). This conversion
is done internally so begin, end and strides receive x, -3, and -1.
The appropriate begin_mask bit is set to indicate the start range is the
full range (ignoring the x).</li><li><code>:</code> indicates that the entire contents of the corresponding dimension
is selected. This is equivalent to `::` or `0::1`. begin, end, and strides
receive 0, 0, and 1, respectively. The appropriate bits in <code>begin_mask</code> and
<code>end_mask</code> are also set.</li></ol><ul><li>Requirements*:
`0 != strides[i] for i in [0, m)`
`ellipsis_mask must be a power of two (only one ellipsis)`</li></ul></div></div><divclass="top"><pclass="src"><aname="v:slice"class="def">slice</a><ahref="src/TensorFlow-GenOps-Core.html#slice"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> index, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` index)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 index</td><tdclass="doc"><p><strong>begin</strong>: begin[i] specifies the offset into the <code>i</code>th dimension of
<code>input</code> to slice from.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 index</td><tdclass="doc"><p><strong>size</strong>: size[i] specifies the number of elements of the <code>i</code>th dimension
of <code>input</code> to slice. If size[i] is -1, all remaining elements in dimension
i are included in the slice (i.e. this is equivalent to setting
size[i] = input.dim_size(i) - begin[i]).</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Return a slice from <code>input</code>.</p><p>The output tensor is a tensor with dimensions described by <code><ahref="TensorFlow-GenOps-Core.html#v:size">size</a></code>
whose values are extracted from <code>input</code> starting at the offsets in
<code>begin</code>.</p><ul><li>Requirements*:
0 <= begin[i] <= begin[i] + size[i] <= Di for i in [0, n)</li></ul></div></div><divclass="top"><pclass="src"><aname="v:quantizedConv2D"class="def">quantizedConv2D</a><ahref="src/TensorFlow-GenOps-Core.html#quantizedConv2D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tinput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` tinput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tfilter, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` tfilter, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` out_type)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 tinput</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tfilter</td><tdclass="doc"><p><strong>filter</strong>: filter's input_depth dimension must match input's depth dimensions.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>min_input</strong>: The float value that the lowest quantized input value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>max_input</strong>: The float value that the highest quantized input value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>min_filter</strong>: The float value that the lowest quantized filter value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>max_filter</strong>: The float value that the highest quantized filter value represents.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>output</strong>, <strong>min_output</strong>, <strong>max_output</strong>)</p><ul><li><strong>output</strong></li><li><strong>min_output</strong>: The float value that the lowest quantized output value represents.</li><li><strong>max_outp
number of the associated minimum, and the highest represents the maximum.
This means that you can only interpret the quantized output in the same way, by
taking the returned minimum and maximum values into account.</p></div></div><divclass="top"><pclass="src"><aname="v:relu6Grad"class="def">relu6Grad</a><ahref="src/TensorFlow-GenOps-Core.html#relu6Grad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>gradients</strong>: The backpropagated gradients to the corresponding Relu6 operation.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>features</strong>: The features passed as input to the corresponding Relu6 operation.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>backprops</strong>: The gradients:
`gradients * (features > 0) * (features < 6)`.</p></td></tr></table></div><divclass="doc"><p>Computes rectified linear 6 gradients for a Relu6 operation.</p></div></div><divclass="top"><pclass="src"><aname="v:avgPoolGrad"class="def">avgPoolGrad</a><ahref="src/TensorFlow-GenOps-Core.html#avgPoolGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>orig_input_shape</strong>: 1-D. Shape of the original input to <code>avg_pool</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>grad</strong>: 4-D with shape `[batch, height, width, channels]`. Gradients w.r.t.
the output of <code>avg_pool</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 4-D. Gradients w.r.t. the input of <code>avg_pool</code>.</p></td></tr></table></div><divclass="doc"><p>Computes gradients of the average pooling function.</p></div></div><divclass="top"><pclass="src"><aname="v:stringSplit"class="def">stringSplit</a><ahref="src/TensorFlow-GenOps-Core.html#stringSplit"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>input</strong>: 1-D. Strings to split.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>delimiter</strong>: 0-D. Delimiter character, or empty string.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p>(<strong>indices</strong>, <strong>values</strong>, <strong>shape</strong>)</p><ul><li><strong>indices</strong>: A dense matrix of int64 representing the indices of the sparse tensor.</li><li><strong>values</strong>: A vector of strings corresponding to the splited values.</li><li><strong>shape</strong>: a length-2 vector of int64 representing the shape of the sparse
tensor, where the first value is N and the second value is the maximum number
of tokens in a single input entry.</li></ul></td></tr></table></div><divclass="doc"><p>Split elements of <code>input</code> based on <code>delimiter</code> into a <code>SparseTensor</code>.</p><p>Let N be the size of source (typically N will be the batch size). Split each
element of <code>input</code> based on <code>delimiter</code> and return a <code>SparseTensor</code>
containing the splitted tokens. Empty tokens are ignored.</p><p><code>delimiter</code> can be empty or a single-byte character. If <code>delimiter</code> is an empty
string, each element of <code>input</code> is split into individual single-byte character
strings, including splitting of UTF-8 multibyte sequences.</p><p>For example:
N = 2, input[0] is 'hello world' and input[1] is 'a b c', then the output
will be</p><p>indices = [0, 0;
0, 1;
1, 0;
1, 1;
1, 2]
shape = [2, 3]
values = [<code>hello</code>, <code>world</code>, <code>a</code>, <code>b</code>, <code>c</code>]</p></div></div><divclass="top"><pclass="src"><aname="v:rank"class="def">rank</a><ahref="src/TensorFlow-GenOps-Core.html#rank"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Returns the rank of a tensor.</p><p>This operation returns an integer representing the rank of <code>input</code>.</p><p>For example:</p><p>```prettyprint
```</p><ul><li>*Note**: The rank of a tensor is not the same as the rank of a matrix. The rank
of a tensor is the number of indices required to uniquely select each element
of the tensor. Rank is also known as "order", "degree", or "ndims."</li></ul></div></div><divclass="top"><pclass="src"><aname="v:reciprocal"class="def">reciprocal</a><ahref="src/TensorFlow-GenOps-Core.html#reciprocal"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes the reciprocal of x element-wise.</p><p>I.e., \(y = 1 / x\).</p></div></div><divclass="top"><pclass="src"><aname="v:reverseSequence"class="def">reverseSequence</a><ahref="src/TensorFlow-GenOps-Core.html#reverseSequence"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tlen, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tlen)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>seq_dim</strong>: The dimension which is partially reversed.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: The input to reverse.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tlen</td><tdclass="doc"><p><strong>seq_lengths</strong>: 1-D with length `input.dims(batch_dim)` and
`max(seq_lengths) < input.dims(seq_dim)`</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The partially reversed input. It has the same shape as <code>input</code>.</p></td></tr></table></div><divclass="doc"><p>Reverses variable length slices.</p><p>This op first slices <code>input</code> along the dimension <code>batch_dim</code>, and for each
slice <code>i</code>, reverses the first `seq_lengths[i]` elements along
the dimension <code>seq_dim</code>.</p><p>The elements of <code>seq_lengths</code> must obey `seq_lengths[i] < input.dims[seq_dim]`,
and <code>seq_lengths</code> must be a vector of length `input.dims[batch_dim]`.</p><p>The output slice <code>i</code> along dimension <code>batch_dim</code> is then given by input
slice <code>i</code>, with the first `seq_lengths[i]` slices along dimension
```</p></div></div><divclass="top"><pclass="src"><aname="v:biasAddGrad"class="def">biasAddGrad</a><ahref="src/TensorFlow-GenOps-Core.html#biasAddGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>out_backprop</strong>: Any number of dimensions.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 1-D with size the feature dimension of <code>out_backprop</code>.</p></td></tr></table></div><divclass="doc"><p>The backward operation for <ahref="BiasAdd.html">BiasAdd</a> on the "bias" tensor.</p><p>It accumulates all the values from out_backprop into the feature dimension.
For NHWC data format, the feature dimension is the last. For NCHW data format,
the feature dimension is the third-to-last.</p></div></div><divclass="top"><pclass="src"><aname="v:addSparseToTensorsMap"class="def">addSparseToTensorsMap</a><ahref="src/TensorFlow-GenOps-Core.html#addSparseToTensorsMap"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sparse_indices</strong>: 2-D. The <code>indices</code> of the <code>SparseTensor</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>sparse_values</strong>: 1-D. The <code>values</code> of the <code>SparseTensor</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sparse_shape</strong>: 1-D. The <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> of the <code>SparseTensor</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p><strong>sparse_handle</strong>: 0-D. The handle of the <code>SparseTensor</code> now stored in the
<code>SparseTensorsMap</code>.</p></td></tr></table></div><divclass="doc"><p>Add a <code>SparseTensor</code> to a <code>SparseTensorsMap</code> return its handle.</p><p>A <code>SparseTensor</code> is represented by three tensors: <code>sparse_indices</code>,
<code>sparse_values</code>, and <code>sparse_shape</code>.</p><p>This operator takes the given <code>SparseTensor</code> and adds it to a container
object (a <code>SparseTensorsMap</code>). A unique key within this container is generated
in the form of an <code>int64</code>, and this is the value that is returned.</p><p>The <code>SparseTensor</code> can then be read out as part of a minibatch by passing
the key as a vector element to <code>TakeManySparseFromTensorsMap</code>. To ensure
the correct <code>SparseTensorsMap</code> is accessed, ensure that the same
<code>container</code> and <code>shared_name</code> are passed to that Op. If no <code>shared_name</code>
is provided here, instead use the *name* of the Operation created by calling
<code>AddSparseToTensorsMap</code> as the <code>shared_name</code> passed to
<code>TakeManySparseFromTensorsMap</code>. Ensure the Operations are colocated.</p></div></div><divclass="top"><pclass="src"><aname="v:tan"class="def">tan</a><ahref="src/TensorFlow-GenOps-Core.html#tan"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes tan of x element-wise.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseReduceSumSparse"class="def">sparseReduceSumSparse</a><ahref="src/TensorFlow-GenOps-Core.html#sparseReduceSumSparse"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>input_indices</strong>: 2-D. `N x R` matrix with the indices of non-empty values in a
SparseTensor, possibly not in canonical ordering.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>input_values</strong>: 1-D. <code>N</code> non-empty values corresponding to <code>input_indices</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>input_shape</strong>: 1-D. Shape of the input SparseTensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>reduction_axes</strong>: 1-D. Length-<code>K</code> vector containing the reduction axes.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p>(<strong>output_indices</strong>, <strong>output_values</strong>, <strong>output_shape</strong>)</p><ul><li><strong>output_indices</strong></li><li><strong>output_values</strong></li><li><strong>output_shape</strong></li></ul></td></tr></table></div><divclass="doc"><p>Computes the sum of elements across dimensions of a SparseTensor.</p><p>This Op takes a SparseTensor and is the sparse counterpart to
`tf.reduce_sum()`. In contrast to SparseReduceSum, this Op returns a
SparseTensor.</p><p>Reduces <code>sp_input</code> along the dimensions given in <code>reduction_axes</code>. Unless
<code>keep_dims</code> is true, the rank of the tensor is reduced by 1 for each entry in
<code>reduction_axes</code>. If <code>keep_dims</code> is true, the reduced dimensions are retained
with length 1.</p><p>If <code>reduction_axes</code> has no entries, all dimensions are reduced, and a tensor
with a single element is returned. Additionally, the axes can be negative,
which are interpreted according to the indexing rules in Python.</p></div></div><divclass="top"><pclass="src"><aname="v:shapeN"class="def">shapeN</a><ahref="src/TensorFlow-GenOps-Core.html#shapeN"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` out_type)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t]</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_type]</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Returns shape of tensors.</p><p>This operation returns N 1-D integer tensors representing shape of `input[i]s`.</p></div></div><divclass="top"><pclass="src"><aname="v:shape"class="def">shape</a><ahref="src/TensorFlow-GenOps-Core.html#shape"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` out_type)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_type</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Returns the shape of a tensor.</p><p>This operation returns a 1-D integer tensor representing the shape of <code>input</code>.</p><p>For example:</p><p>```prettyprint
```</p></div></div><divclass="top"><pclass="src"><aname="v:unique"class="def">unique</a><ahref="src/TensorFlow-GenOps-Core.html#unique"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> out_idx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` out_idx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong>: 1-D.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_idx)</td><tdclass="doc"><p>(<strong>y</strong>, <strong>idx</strong>)</p><ul><li><strong>y</strong>: 1-D.</li><li><strong>idx</strong>: 1-D.</li></ul></td></tr></table></div><divclass="doc"><p>Finds unique elements in a 1-D tensor.</p><p>This operation returns a tensor <code>y</code> containing all of the unique elements of <code>x</code>
sorted in the same order that they occur in <code>x</code>. This operation also returns a
tensor <code>idx</code> the same size as <code>x</code> that contains the index of each value of <code>x</code>
in the unique output <code>y</code>. In other words:</p><p>`y[idx[i]] = x[i] for i in [0, 1,...,rank(x) - 1]`</p><p>For example:</p><p>```prettyprint
```</p></div></div><divclass="top"><pclass="src"><aname="v:truncatedNormal"class="def">truncatedNormal</a><ahref="src/TensorFlow-GenOps-Core.html#truncatedNormal"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>shape</strong>: The shape of the output tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype)</td><tdclass="doc"><p><strong>output</strong>: A tensor of the specified shape filled with random truncated normal
values.</p></td></tr></table></div><divclass="doc"><p>Outputs random values from a truncated normal distribution.</p><p>The generated values follow a normal distribution with mean 0 and standard
deviation 1, except that values whose magnitude is more than 2 standard
deviations from the mean are dropped and re-picked.</p></div></div><divclass="top"><pclass="src"><aname="v:invertPermutation"class="def">invertPermutation</a><ahref="src/TensorFlow-GenOps-Core.html#invertPermutation"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong>: 1-D.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong>: 1-D.</p></td></tr></table></div><divclass="doc"><p>Computes the inverse permutation of a tensor.</p><p>This operation computes the inverse of an index permutation. It takes a 1-D
integer tensor <code>x</code>, which represents the indices of a zero-based array, and
swaps each value with its index position. In other words, for an output tensor
<code>y</code> and an input tensor <code>x</code>, this operation computes the following:</p><p>`y[x[i]] = i for i in [0, 1, ..., len(x) - 1]`</p><p>The values must include 0. There can be no duplicate values or negative values.</p><p>For example:</p><p>```prettyprint
# tensor <code>x</code> is [3, 4, 0, 2, 1]
invert_permutation(x) ==> [2, 4, 3, 0, 1]
```</p></div></div><divclass="top"><pclass="src"><aname="v:checkNumerics"class="def">checkNumerics</a><ahref="src/TensorFlow-GenOps-Core.html#checkNumerics"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>tensor</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Checks a tensor for NaN and Inf values.</p><p>When run, reports an <code>InvalidArgument</code> error if <code>tensor</code> has any values
that are not a number (NaN) or infinity (Inf). Otherwise, passes <code>tensor</code> as-is.</p></div></div><divclass="top"><pclass="src"><aname="v:uniformCandidateSampler"class="def">uniformCandidateSampler</a><ahref="src/TensorFlow-GenOps-Core.html#uniformCandidateSampler"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_sampled</strong>: Number of candidates to randomly sample per batch.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_true</strong>: Number of true labels per context.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>range_max</strong>: The sampler will sample integers from the interval [0, range_max).</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>unique</strong>: If unique is true, we sample with rejection, so that all sampled
candidates in a batch are unique. This requires some approximation to
estimate the post-rejection sampling probabilities.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>true_classes</strong>: A batch_size * num_true matrix, in which each row contains the
IDs of the num_true target_classes in the corresponding original label.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>sampled_candidates</strong>, <strong>true_expected_count</strong>, <strong>sampled_expected_count</strong>)</p><ul><li><strong>sampled_candidates</strong>: A vector of length num_sampled, in which each element is
the ID of a sampled candidate.</li><li><strong>true_expected_count</strong>: A batch_size * num_true matrix, representing
the number of times each candidate is expected to occur in a batch
of sampled candidates. If unique=true, then this is a probability.</li><li><strong>sampled_expected_count</strong>: A vector of length num_sampled, for each sampled
candidate representing the number of times the candidate is expected
to occur in a batch of sampled candidates. If unique=true, then this is a
probability.</li></ul></td></tr></table></div><divclass="doc"><p>Generates labels for candidate sampling with a uniform distribution.</p><p>See explanations of candidate sampling and the data formats at
go/candidate-sampling.</p><p>For each batch, this op picks a single set of sampled candidate labels.</p><p>The advantages of sampling candidates per-batch are simplicity and the
possibility of efficient dense matrix multiplication. The disadvantage is that
the sampled candidates must be chosen independently of the context and of the
true labels.</p></div></div><divclass="top"><pclass="src"><aname="v:gather"class="def">gather</a><ahref="src/TensorFlow-GenOps-Core.html#gather"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tparams, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 tparams</td><tdclass="doc"><p><strong>params</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>indices</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> tparams</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Gather slices from <code>params</code> according to <code>indices</code>.</p><p><code>indices</code> must be an integer tensor of any dimension (usually 0-D or 1-D).
Produces an output tensor with shape `indices.shape + params.shape[1:]` where:</p><p>```python
# Scalar indices
output[:, ..., :] = params[indices, :, ... :]</p><p># Vector indices
```</p><p>If <code>indices</code> is a permutation and `len(indices) == params.shape[0]` then
this operation will permute <code>params</code> accordingly.</p><p><ahref="div">style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;"</a>
<ahref="/div">/div</a></p></div></div><divclass="top"><pclass="src"><aname="v:const"class="def">const</a><ahref="src/TensorFlow-GenOps-Core.html#const"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Returns a constant tensor.</p></div></div><divclass="top"><pclass="src"><aname="v:fill"class="def">fill</a><ahref="src/TensorFlow-GenOps-Core.html#fill"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>dims</strong>: 1-D. Represents the shape of the output tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>value</strong>: 0-D (scalar). Value to fill the returned tensor.</p><p><code>compatibility(numpy)
Equivalent to np.full
</code>end_compatibility</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Creates a tensor filled with a scalar value.</p><p>This operation creates a tensor of shape <code>dims</code> and fills it with <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>.</p><p>For example:</p><p>```prettyprint
# Output tensor has shape [2, 3].
fill([2, 3], 9) ==> [[9, 9, 9]
[9, 9, 9]]
```</p></div></div><divclass="top"><pclass="src"><aname="v:editDistance"class="def">editDistance</a><ahref="src/TensorFlow-GenOps-Core.html#editDistance"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>hypothesis_indices</strong>: The indices of the hypothesis list SparseTensor.
This is an N x R int64 matrix.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>hypothesis_values</strong>: The values of the hypothesis list SparseTensor.
This is an N-length vector.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>hypothesis_shape</strong>: The shape of the hypothesis list SparseTensor.
This is an R-length vector.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>truth_indices</strong>: The indices of the truth list SparseTensor.
This is an M x R int64 matrix.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>truth_values</strong>: The values of the truth list SparseTensor.
This is an M-length vector.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>truth_shape</strong>: truth indices, vector.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>output</strong>: A dense float tensor with rank R - 1.</p><p>For the example input:</p><p>// hypothesis represents a 2x1 matrix with variable-length values:
(truth_indices, truth_values, truth_shape).</p><p>The inputs are:</p></div></div><divclass="top"><pclass="src"><aname="v:reverse"class="def">reverse</a><ahref="src/TensorFlow-GenOps-Core.html#reverse"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>tensor</strong>: Up to 8-D.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>dims</strong>: 1-D. The dimensions to reverse.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The same shape as <code>tensor</code>.</p></td></tr></table></div><divclass="doc"><p>Reverses specific dimensions of a tensor.</p><p>Given a <code>tensor</code>, and a <code>bool</code> tensor <code>dims</code> representing the dimensions
of <code>tensor</code>, this operation reverses each dimension i of <code>tensor</code> where
`dims[i]` is <code><ahref="../base-4.8.2.0/Data-Bool.html#v:True">True</a></code>.</p><p><code>tensor</code> can have up to 8 dimensions. The number of dimensions
of <code>tensor</code> must equal the number of elements in <code>dims</code>. In other words:</p><p>`rank(tensor) = size(dims)`</p><p>For example:</p><p>```prettyprint
# tensor <code>t</code> is [[[[ 0, 1, 2, 3],
# [ 4, 5, 6, 7],
# [ 8, 9, 10, 11]],
# [[12, 13, 14, 15],
# [16, 17, 18, 19],
# [20, 21, 22, 23]]]]
# tensor <code>t</code> shape is [1, 2, 3, 4]</p><p># <code>dims</code> is [False, False, False, True]
reverse(t, dims) ==> [[[[ 3, 2, 1, 0],
[ 7, 6, 5, 4],
[ 11, 10, 9, 8]],
[[15, 14, 13, 12],
[19, 18, 17, 16],
[23, 22, 21, 20]]]]</p><p># <code>dims</code> is [False, True, False, False]
```</p></div></div><divclass="top"><pclass="src"><aname="v:matrixSetDiag"class="def">matrixSetDiag</a><ahref="src/TensorFlow-GenOps-Core.html#matrixSetDiag"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Rank `k+1`, where `k >= 1`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>diagonal</strong>: Rank <code>k</code>, where `k >= 1`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Rank `k+1`, with `output.shape = input.shape`.</p></td></tr></table></div><divclass="doc"><p>Returns a batched matrix tensor with new batched diagonal values.</p><p>Given <code>input</code> and <code>diagonal</code>, this operation returns a tensor with the
same shape and values as <code>input</code>, except for the main diagonal of the
innermost matrices. These will be overwritten by the values in <code>diagonal</code>.</p><p>The output is computed as follows:</p><p>Assume <code>input</code> has `k+1` dimensions `[I, J, K, ..., M, N]` and <code>diagonal</code> has
<code>k</code> dimensions `[I, J, K, ..., min(M, N)]`. Then the output is a
tensor of rank `k+1` with dimensions `[I, J, K, ..., M, N]` where:</p><ul><li>`output[i, j, k, ..., m, n] = diagonal[i, j, k, ..., n]` for `m == n`.</li><li>`output[i, j, k, ..., m, n] = input[i, j, k, ..., m, n]` for `m != n`.</li></ul></div></div><divclass="top"><pclass="src"><aname="v:matrixDiag"class="def">matrixDiag</a><ahref="src/TensorFlow-GenOps-Core.html#matrixDiag"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>diagonal</strong>: Rank <code>k</code>, where `k >= 1`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Rank `k+1`, with `output.shape = diagonal.shape + [diagonal.shape[-1]]`.</p></td></tr></table></div><divclass="doc"><p>Returns a batched diagonal tensor with a given batched diagonal values.</p><p>Given a <code>diagonal</code>, this operation returns a tensor with the <code>diagonal</code> and
everything else padded with zeros. The diagonal is computed as follows:</p><p>Assume <code>diagonal</code> has <code>k</code> dimensions `[I, J, K, ..., N]`, then the output is a
tensor of rank `k+1` with dimensions [I, J, K, ..., N, N]` where:</p><p>`output[i, j, k, ..., m, n] = 1{m=n} * diagonal[i, j, k, ..., n]`.</p><p>For example:</p><p>```prettyprint
```</p></div></div><divclass="top"><pclass="src"><aname="v:diag"class="def">diag</a><ahref="src/TensorFlow-GenOps-Core.html#diag"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>diagonal</strong>: Rank k tensor where k is at most 3.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Returns a diagonal tensor with a given diagonal values.</p><p>Given a <code>diagonal</code>, this operation returns a tensor with the <code>diagonal</code> and
everything else padded with zeros. The diagonal is computed as follows:</p><p>Assume <code>diagonal</code> has dimensions [D1,..., Dk], then the output is a tensor of
```</p></div></div><divclass="top"><pclass="src"><aname="v:immutableConst"class="def">immutableConst</a><ahref="src/TensorFlow-GenOps-Core.html#immutableConst"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:Shape">Shape</a></td><tdclass="doc"><p><strong>shape</strong>: Shape of the returned tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype</td><tdclass="doc"><p><strong>tensor</strong></p></td></tr></table></div><divclass="doc"><p>Returns immutable tensor from memory region.</p><p>The current implementation memmaps the tensor from a file.</p></div></div><divclass="top"><pclass="src"><aname="v:concat"class="def">concat</a><ahref="src/TensorFlow-GenOps-Core.html#concat"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>concat_dim</strong>: 0-D. The dimension along which to concatenate. Must be in the
range [0, rank(values)).</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t]</td><tdclass="doc"><p><strong>values</strong>: The <code>N</code> Tensors to concatenate. Their ranks and types must match,
and their sizes must match in all dimensions except <code>concat_dim</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: A <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> with the concatenation of values stacked along the
<code>concat_dim</code> dimension. This tensor's shape matches that of <code>values</code> except
in <code>concat_dim</code> where it has the sum of the sizes.</p></td></tr></table></div><divclass="doc"><p>Concatenates tensors along one dimension.</p></div></div><divclass="top"><pclass="src"><aname="v:unpack"class="def">unpack</a><ahref="src/TensorFlow-GenOps-Core.html#unpack"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>value</strong>: 1-D or higher, with <code>axis</code> dimension size equal to <code>num</code>.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t]</td><tdclass="doc"><p><strong>output</strong>: The list of tensors unpacked from <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>.</p></td></tr></table></div><divclass="doc"><p>Unpacks a given dimension of a rank-<code>R</code> tensor into <code>num</code> rank-`(R-1)` tensors.</p><p>Unpacks <code>num</code> tensors from <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code> by chipping it along the <code>axis</code> dimension.
For example, given a tensor of shape `(A, B, C, D)`;</p><p>If `axis == 0` then the i'th tensor in <code>output</code> is the slice `value[i, :, :, :]`
and each tensor in <code>output</code> will have shape `(B, C, D)`. (Note that the
dimension unpacked along is gone, unlike <code><ahref="TensorFlow-GenOps-Core.html#v:split">split</a></code>).</p><p>If `axis == 1` then the i'th tensor in <code>output</code> is the slice `value[:, i, :, :]`
and each tensor in <code>output</code> will have shape `(A, C, D)`.
Etc.</p><p>This is the opposite of <code><ahref="TensorFlow-GenOps-Core.html#v:pack">pack</a></code>.</p></div></div><divclass="top"><pclass="src"><aname="v:fact"class="def">fact</a><ahref="src/TensorFlow-GenOps-Core.html#fact"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>fact</strong></p></td></tr></table></div><divclass="doc"><p>Output a fact about factorials.</p></div></div><divclass="top"><pclass="src"><aname="v:abs"class="def">abs</a><ahref="src/TensorFlow-GenOps-Core.html#abs"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes the absolute value of a tensor.</p><p>Given a tensor <code>x</code>, this operation returns a tensor containing the absolute
value of each element in <code>x</code>. For example, if x is an input element and y is
an output element, this operation computes \(y = |x|\).</p></div></div><divclass="top"><pclass="src"><aname="v:softmax"class="def">softmax</a><ahref="src/TensorFlow-GenOps-Core.html#softmax"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>logits</strong>: 2-D with shape `[batch_size, num_classes]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>softmax</strong>: Same shape as <code>logits</code>.</p></td></tr></table></div><divclass="doc"><p>Computes softmax activations.</p><p>For each batch <code>i</code> and class <code>j</code> we have</p><p>softmax[i, j] = exp(logits[i, j]) / sum_j(exp(logits[i, j]))</p></div></div><divclass="top"><pclass="src"><aname="v:reverseV2"class="def">reverseV2</a><ahref="src/TensorFlow-GenOps-Core.html#reverseV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>tensor</strong>: Up to 8-D.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>axis</strong>: 1-D. The indices of the dimensions to reverse.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The same shape as <code>tensor</code>.</p></td></tr></table></div><divclass="doc"><p>Reverses specific dimensions of a tensor.</p><p>Given a <code>tensor</code>, and a <code>int32</code> tensor <code>axis</code> representing the set of
dimensions of <code>tensor</code> to reverse. This operation reverses each dimension
<code>i</code> for which there exists <code>j</code> s.t. `axis[j] == i`.</p><p><code>tensor</code> can have up to 8 dimensions. The number of dimensions specified
in <code>axis</code> may be 0 or more entries. If an index is specified more than
once, a InvalidArgument error is raised.</p><p>For example:</p><p>```prettyprint
# tensor <code>t</code> is [[[[ 0, 1, 2, 3],
# [ 4, 5, 6, 7],
# [ 8, 9, 10, 11]],
# [[12, 13, 14, 15],
# [16, 17, 18, 19],
# [20, 21, 22, 23]]]]
# tensor <code>t</code> shape is [1, 2, 3, 4]</p><p># <code>dims</code> is [3] or <code>dims</code> is -1
reverse(t, dims) ==> [[[[ 3, 2, 1, 0],
[ 7, 6, 5, 4],
[ 11, 10, 9, 8]],
[[15, 14, 13, 12],
[19, 18, 17, 16],
[23, 22, 21, 20]]]]</p><p># <code>dims</code> is '[1]' (or <code>dims</code> is '[-3]')
reverse(t, dims) ==> [[[[12, 13, 14, 15],
[16, 17, 18, 19],
[20, 21, 22, 23]
[[ 0, 1, 2, 3],
[ 4, 5, 6, 7],
[ 8, 9, 10, 11]]]]</p><p># <code>dims</code> is '[2]' (or <code>dims</code> is '[-2]')
reverse(t, dims) ==> [[[[8, 9, 10, 11],
[4, 5, 6, 7],
[0, 1, 2, 3]]
[[20, 21, 22, 23],
[16, 17, 18, 19],
[12, 13, 14, 15]]]]
```</p></div></div><divclass="top"><pclass="src"><aname="v:identity"class="def">identity</a><ahref="src/TensorFlow-GenOps-Core.html#identity"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Return a tensor with the same shape and contents as the input tensor or value.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseAdd"class="def">sparseAdd</a><ahref="src/TensorFlow-GenOps-Core.html#sparseAdd"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> treal, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` treal)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>a_indices</strong>: 2-D. The <code>indices</code> of the first <code>SparseTensor</code>, size `[nnz, ndims]` Matrix.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>a_values</strong>: 1-D. The <code>values</code> of the first <code>SparseTensor</code>, size `[nnz]` Vector.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>a_shape</strong>: 1-D. The <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> of the first <code>SparseTensor</code>, size `[ndims]` Vector.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>b_indices</strong>: 2-D. The <code>indices</code> of the second <code>SparseTensor</code>, size `[nnz, ndims]` Matrix.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tenso
pair takes space.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p>(<strong>sum_indices</strong>, <strong>sum_values</strong>, <strong>sum_shape</strong>)</p><ul><li><strong>sum_indices</strong></li><li><strong>sum_values</strong></li><li><strong>sum_shape</strong></li></ul></td></tr></table></div><divclass="doc"><p>Adds two <code>SparseTensor</code> objects to produce another <code>SparseTensor</code>.</p><p>The input <code>SparseTensor</code> objects' indices are assumed ordered in standard
lexicographic order. If this is not the case, before this step run
<code>SparseReorder</code> to restore index ordering.</p><p>By default, if two values sum to zero at some index, the output <code>SparseTensor</code>
would still include that particular location in its index, storing a zero in the
corresponding value slot. To override this, callers can specify <code>thresh</code>,
indicating that if the sum has a magnitude strictly smaller than <code>thresh</code>, its
corresponding value and index would then not be included. In particular,
`thresh == 0` (default) means everything is kept and actual thresholding happens
only for a positive value.</p><p>In the following shapes, <code>nnz</code> is the count after taking <code>thresh</code> into account.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseApplyCenteredRMSProp"class="def">sparseApplyCenteredRMSProp</a><ahref="src/TensorFlow-GenOps-Core.html#sparseApplyCenteredRMSProp"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>mg</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ms</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>mom</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>lr</strong>: Scaling factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 t</td><tdclass="doc"><p><strong>rho</strong>: Decay rate. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v7 t</td><tdclass="doc"><p><strong>momentum</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v8 t</td><tdclass="doc"><p><strong>epsilon</strong>: Ridge term. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v9 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v10 tindices</td><tdclass="doc"><p><strong>indices</strong>: A vector of indices into the first dimension of var, ms and mom.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/Ten
(i.e., the variance) for normalization, as opposed to regular RMSProp, which
uses the (uncentered) second moment. This often helps with training, but is
slightly more expensive in terms of computation and memory.</p><p>Note that in dense implementation of this algorithm, mg, ms, and mom will
update even if the grad is zero, but in this sparse implementation, mg, ms,
and mom will not update in iterations during which the grad is zero.</p><p>mean_square = decay * mean_square + (1-decay) * gradient ** 2
mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms + epsilon)
var <- var - mom</p></div></div><divclass="top"><pclass="src"><aname="v:addN"class="def">addN</a><ahref="src/TensorFlow-GenOps-Core.html#addN"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t]</td><tdclass="doc"><p><strong>inputs</strong>: Must all be the same size and shape.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>sum</strong></p></td></tr></table></div><divclass="doc"><p>Add all input tensors element wise.</p></div></div><divclass="top"><pclass="src"><aname="v:concatOffset"class="def">concatOffset</a><ahref="src/TensorFlow-GenOps-Core.html#concatOffset"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>concat_dim</strong>: The dimension along which to concatenate.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>]</td><tdclass="doc"><p><strong>shape</strong>: The <code>N</code> int32 vectors representing shape of tensors being concatenated.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>]</td><tdclass="doc"><p><strong>offset</strong>: The <code>N</code> int32 vectors representing the starting offset
of input tensors within the concatenated output.</p><p>This is typically used by gradient computations for a concat operation.</p></td></tr></table></div><divclass="doc"><p>Computes offsets of concat inputs within its output.</p><p>For example:</p><p>```prettyprint
```</p></div></div><divclass="top"><pclass="src"><aname="v:concatV2"class="def">concatV2</a><ahref="src/TensorFlow-GenOps-Core.html#concatV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t]</td><tdclass="doc"><p><strong>values</strong>: List of <code>N</code> Tensors to concatenate. Their ranks and types must match,
and their sizes must match in all dimensions except <code>concat_dim</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>axis</strong>: 0-D. The dimension along which to concatenate. Must be in the
range [0, rank(values)).</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: A <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> with the concatenation of values stacked along the
<code>concat_dim</code> dimension. This tensor's shape matches that of <code>values</code> except
in <code>concat_dim</code> where it has the sum of the sizes.</p></td></tr></table></div><divclass="doc"><p>Concatenates tensors along one dimension.</p></div></div><divclass="top"><pclass="src"><aname="v:zerosLike"class="def">zerosLike</a><ahref="src/TensorFlow-GenOps-Core.html#zerosLike"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong>: a tensor of type T.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong>: a tensor of the same shape and type as x but filled with zeros.</p></td></tr></table></div><divclass="doc"><p>Returns a tensor of zeros with the same shape and type as x.</p></div></div><divclass="top"><pclass="src"><aname="v:applyCenteredRMSProp"class="def">applyCenteredRMSProp</a><ahref="src/TensorFlow-GenOps-Core.html#applyCenteredRMSProp"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>mg</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ms</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>mom</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>lr</strong>: Scaling factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 t</td><tdclass="doc"><p><strong>rho</strong>: Decay rate. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v7 t</td><tdclass="doc"><p><strong>momentum</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v8 t</td><tdclass="doc"><p><strong>epsilon</strong>: Ridge
(i.e., the variance) for normalization, as opposed to regular RMSProp, which
uses the (uncentered) second moment. This often helps with training, but is
slightly more expensive in terms of computation and memory.</p><p>Note that in dense implementation of this algorithm, mg, ms, and mom will
update even if the grad is zero, but in this sparse implementation, mg, ms,
and mom will not update in iterations during which the grad is zero.</p><p>mean_square = decay * mean_square + (1-decay) * gradient ** 2
mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms - mg * mg + epsilon)
var <- var - mom</p></div></div><divclass="top"><pclass="src"><aname="v:applyRMSProp"class="def">applyRMSProp</a><ahref="src/TensorFlow-GenOps-Core.html#applyRMSProp"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ms</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>mom</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>lr</strong>: Scaling factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>rho</strong>: Decay rate. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 t</td><tdclass="doc"><p><strong>momentum</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v7 t</td><tdclass="doc"><p><strong>epsilon</strong>: Ridge term. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v8 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>out</strong>: Same as "var".</p></td></tr></table></div><divclass="doc"><p>Update '*var' according to the RMSProp algorithm.</p><p>Note that in dense implementation of this algorithm, ms and mom will
update even if the grad is zero, but in this sparse implementation, ms
and mom will not update in iterations during which the grad is zero.</p><p>mean_square = decay * mean_square + (1-decay) * gradient ** 2
Delta = learning_rate * gradient / sqrt(mean_square + epsilon)</p><p>ms <- rho * ms_{t-1} + (1-rho) * grad * grad
mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms + epsilon)
var <- var - mom</p></div></div><divclass="top"><pclass="src"><aname="v:assignAddVariableOp"class="def">assignAddVariableOp</a><ahref="src/TensorFlow-GenOps-Core.html#assignAddVariableOp"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ResourceHandle">ResourceHandle</a> dtype</td><tdclass="doc"><p><strong>resource</strong>: handle to the resource in which to store the variable.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 dtype</td><tdclass="doc"><p><strong>value</strong>: the value by which the variable will be incremented.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Adds a value to the current value of a variable.</p><p>Any ReadVariableOp which depends directly or indirectly on this assign is
guaranteed to see the incremented value or a subsequent newer one.</p><p>Outputs the incremented value, which can be used to totally order the
increments to this variable.</p></div></div><divclass="top"><pclass="src"><aname="v:applyAdam"class="def">applyAdam</a><ahref="src/TensorFlow-GenOps-Core.html#applyAdam"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>m</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>v</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>beta1_power</strong>: Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>beta2_power</strong>: Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 t</td><tdclass="doc"><p><strong>lr</strong>: Scaling factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v7 t</td><tdclass="doc"><p><strong>beta1</strong>: Momentum factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v8 t</td><tdclass="doc"><p><strong>beta2</strong>: Momentum factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v9 t</td><tdclass="doc"><p><strong>epsilon</strong>: Ridge term. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v10 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>out</strong>: Same as "var".</p></td></tr></table></div><divclass="doc"><p>Update '*var' according to the Adam algorithm.</p><p>lr_t <- learning_rate * sqrt(1 - beta2^t) / (1 - beta1^t)
variable <- variable - lr_t * m_t / (sqrt(v_t) + epsilon)</p></div></div><divclass="top"><pclass="src"><aname="v:extractGlimpse"class="def">extractGlimpse</a><ahref="src/TensorFlow-GenOps-Core.html#extractGlimpse"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>input</strong>: A 4-D float tensor of shape `[batch_size, height, width, channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>size</strong>: A 1-D tensor of 2 elements containing the size of the glimpses
to extract. The glimpse height must be specified first, following
by the glimpse width.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>offsets</strong>: A 2-D integer tensor of shape `[batch_size, 2]` containing
the x, y locations of the center of each window.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>glimpse</strong>: A tensor representing the glimpses `[batch_size,
glimpse_height, glimpse_width, channels]`.</p></td></tr></table></div><divclass="doc"><p>Extracts a glimpse from the input tensor.</p><p>Returns a set of windows called glimpses extracted at location
<code>offsets</code> from the input tensor. If the windows only partially
overlaps the inputs, the non overlapping areas will be filled with
random noise.</p><p>The result is a 4-D tensor of shape `[batch_size, glimpse_height,
glimpse_width, channels]`. The channels and batch dimensions are the
same as that of the input tensor. The height and width of the output
windows are specified in the <code><ahref="TensorFlow-GenOps-Core.html#v:size">size</a></code> parameter.</p><p>The argument <code>normalized</code> and <code>centered</code> controls how the windows are built:</p><ul><li>If the coordinates are normalized but not centered, 0.0 and 1.0
correspond to the minimum and maximum of each height and width
dimension.</li><li>If the coordinates are both normalized and centered, they range from</li><li>1.0 to 1.0. The coordinates (-1.0, -1.0) correspond to the upper
left corner, the lower right corner is located at (1.0, 1.0) and the
center is at (0, 0).</li><li>If the coordinates are not normalized they are interpreted as
numbers of pixels.</li></ul></div></div><divclass="top"><pclass="src"><aname="v:sparseApplyMomentum"class="def">sparseApplyMomentum</a><ahref="src/TensorFlow-GenOps-Core.html#sparseApplyMomentum"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>accum</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>lr</strong>: Learning rate. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 tindices</td><tdclass="doc"><p><strong>indices</strong>: A vector of indices into the first dimension of var and accum.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 t</td><tdclass="doc"><p><strong>momentum</strong>: Momentum. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>out</strong>: Same as "var".</p></td></tr></table></div><divclass="doc"><p>Update relevant entries in '*var' and '*accum' according to the momentum scheme.</p><p>Set use_nesterov = True if you want to use Nesterov momentum.</p><p>That is for rows we have grad for, we update var and accum as follows:</p><p>accum = accum * momentum + grad
var -= lr * accum</p></div></div><divclass="top"><pclass="src"><aname="v:applyMomentum"class="def">applyMomentum</a><ahref="src/TensorFlow-GenOps-Core.html#applyMomentum"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>accum</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>lr</strong>: Scaling factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>momentum</strong>: Momentum. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>out</strong>: Same as "var".</p></td></tr></table></div><divclass="doc"><p>Update '*var' according to the momentum scheme. Set use_nesterov = True if you</p><p>want to use Nesterov momentum.</p><p>accum = accum * momentum + grad
var -= lr * accum</p></div></div><divclass="top"><pclass="src"><aname="v:fIFOQueue"class="def">fIFOQueue</a><ahref="src/TensorFlow-GenOps-Core.html#fIFOQueue"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>)</td><tdclass="doc"><p><strong>handle</strong>: The handle to the queue.</p></td></tr></table></div><divclass="doc"><p>A queue that produces elements in first-in first-out order.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseApplyFtrl"class="def">sparseApplyFtrl</a><ahref="src/TensorFlow-GenOps-Core.html#sparseApplyFtrl"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>accum</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>linear</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 tindices</td><tdclass="doc"><p><strong>indices</strong>: A vector of indices into the first dimension of var and accum.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 t</td><tdclass="doc"><p><strong>lr</strong>: Scaling factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v7 t</td><tdclass="doc"><p><strong>l1</strong>: L1 regularization. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v8 t</td><tdclass="doc"><p><strong>l2</strong>: L2 regularization. Must be a scalar.</p></td></tr><tr><tdclass="src">-><a
accum_new = accum + grad * grad
linear += grad + (accum_new^(-lr_power) - accum^(-lr_power)) / lr * var
var = (sign(linear) * l1 - linear) / quadratic if |linear| > l1 else 0.0
accum = accum_new</p></div></div><divclass="top"><pclass="src"><aname="v:sparseApplyAdagradDA"class="def">sparseApplyAdagradDA</a><ahref="src/TensorFlow-GenOps-Core.html#sparseApplyAdagradDA"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>gradient_accumulator</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>gradient_squared_accumulator</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 tindices</td><tdclass="doc"><p><strong>indices</strong>: A vector of indices into the first dimension of var and accum.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 t</td><tdclass="doc"><p><strong>lr</strong>: Learning rate. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v7 t</td><tdclass="doc"><p><strong>l1</strong>: L1 regularization. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v8 t</td><tdclass="doc"><p><strong>l2</strong>: L2 regularization. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v9 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>global_step</strong>: Training step number. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>out</strong>: Same as "var".</p></td></tr></table></div><divclass="doc"><p>Update entries in '*var' and '*accum' according to the proximal adagrad scheme.</p></div></
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:applyAdagradDA"class="def">applyAdagradDA</a><ahref="src/TensorFlow-GenOps-Core.html#applyAdagradDA"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>gradient_accumulator</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>gradient_squared_accumulator</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>lr</strong>: Scaling factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 t</td><tdclass="doc"><p><strong>l1</strong>: L1 regularization. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v7 t</td><tdclass="doc"><p><strong>l2</strong>: L2 regularization. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v8 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>global_step</strong>: Training step number. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>out</strong>: Same as "var".</p></td></tr></table></div><divclass="doc"><p>Update '*var' according to the proximal adagrad scheme.</p></div></div><divclass="top"><pclass="src"><aname="v:applyAdagrad"class="def">applyAdagrad</a><ahref="src/TensorFlow-GenOps-Core.html#applyAdagrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-
var -= lr * grad * (1 / sqrt(accum))</p></div></div><divclass="top"><pclass="src"><aname="v:sigmoidGrad"class="def">sigmoidGrad</a><ahref="src/TensorFlow-GenOps-Core.html#sigmoidGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Computes the gradient of the sigmoid of <code>x</code> wrt its input.</p><p>Specifically, `grad = dy * y * (1 - y)`, where `y = sigmoid(x)`, and
<code>dy</code> is the corresponding input gradient.</p></div></div><divclass="top"><pclass="src"><aname="v:applyAdadelta"class="def">applyAdadelta</a><ahref="src/TensorFlow-GenOps-Core.html#applyAdadelta"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>accum</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>accum_update</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>lr</strong>: Scaling factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>rho</strong>: Decay factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 t</td><tdclass="doc"><p><strong>epsilon</strong>: Constant factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v7 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>out</strong>: Same as "var".</p></td></tr></table></div><divclass="doc"><p>Update '*var' according to the adadelta scheme.</p><p>accum = rho() * accum + (1 - rho()) * grad.square();
var -= update;</p></div></div><divclass="top"><pclass="src"><aname="v:sparseApplyProximalGradientDescent"class="def">sparseApplyProximalGradientDescent</a><ahref="src/TensorFlow-GenOps-Core.html#sparseApplyProximalGradientDescent"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>alpha</strong>: Scaling factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>l1</strong>: L1 regularization. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>l2</strong>: L2 regularization. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 tindices</td><tdclass="doc"><p><strong>indices</strong>: A vector of indices into the first dimension of var and accum.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>out</strong>: Same as "var".</p></td></tr></table></div><divclass="doc"><p>Sparse update '*var' as FOBOS algorithm with fixed learning rate.</p><p>That is for rows we have grad for, we update var as follows:
prox_v = var - alpha * grad
var = sign(prox_v)/(1+alpha*l2) * max{|prox_v|-alpha*l1,0}</p></div></div><divclass="top"><pclass="src"><aname="v:applyProximalGradientDescent"class="def">applyProximalGradientDescent</a><ahref="src/TensorFlow-GenOps-Core.html#applyProximalGradientDescent"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>alpha</strong>: Scaling factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>l1</strong>: L1 regularization. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>l2</strong>: L2 regularization. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>delta</strong>: The change.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>out</strong>: Same as "var".</p></td></tr></table></div><divclass="doc"><p>Update '*var' as FOBOS algorithm with fixed learning rate.</p><p>prox_v = var - alpha * delta
var = sign(prox_v)/(1+alpha*l2) * max{|prox_v|-alpha*l1,0}</p></div></div><divclass="top"><pclass="src"><aname="v:matrixSolve"class="def">matrixSolve</a><ahref="src/TensorFlow-GenOps-Core.html#matrixSolve"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>matrix</strong>: Shape is `[..., M, M]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>rhs</strong>: Shape is `[..., M, K]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Shape is `[..., M, K]`.</p></td></tr></table></div><divclass="doc"><p>Solves systems of linear equations.</p><p><code>Matrix</code> is a tensor of shape `[..., M, M]` whose inner-most 2 dimensions
form square matrices. <code>Rhs</code> is a tensor of shape `[..., M, K]`. The <code>output</code> is
a tensor shape `[..., M, K]`. If <code>adjoint</code> is <code><ahref="../base-4.8.2.0/Data-Bool.html#v:False">False</a></code> then each output matrix
If <code>adjoint</code> is <code><ahref="../base-4.8.2.0/Data-Bool.html#v:True">True</a></code> then each output matrix satisfies
`adjoint(matrix[..., :, :]) * output[..., :, :] = rhs[..., :, :]`.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseApplyProximalAdagrad"class="def">sparseApplyProximalAdagrad</a><ahref="src/TensorFlow-GenOps-Core.html#sparseApplyProximalAdagrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>accum</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>lr</strong>: Learning rate. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>l1</strong>: L1 regularization. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>l2</strong>: L2 regularization. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v7 tindices</td><tdclass="doc"><p><strong>indices</strong>: A vector of indices into the first dimension of var and accum.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>out</strong>: Same as "var".</p></td></tr></table></div><divclass="doc"><p>Sparse update entries in '*var' and '*accum' according to FOBOS algorithm.</p><p>That is for rows we have grad for, we update var and accum as follows:
accum += grad * grad
prox_v = var
prox_v -= lr * grad * (1 / sqrt(accum))
var = sign(prox_v)/(1+lr*l2) * max{|prox_v|-lr*l1,0}</p></div></div><divclass="top"><pclass="src"><aname="v:applyGradientDescent"class="def">applyGradientDescent</a><ahref="src/TensorFlow-GenOps-Core.html#applyGradientDescent"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>alpha</strong>: Scaling factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>delta</strong>: The change.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>out</strong>: Same as "var".</p></td></tr></table></div><divclass="doc"><p>Update '*var' by subtracting <code>alpha</code> * <code>delta</code> from it.</p></div></div><divclass="top"><pclass="src"><aname="v:batchNormWithGlobalNormalization"class="def">batchNormWithGlobalNormalization</a><ahref="src/TensorFlow-GenOps-Core.html#batchNormWithGlobalNormalization"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>scale_after_normalization</strong>: A bool indicating whether the resulted tensor
needs to be multiplied with gamma.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>variance_epsilon</strong>: A small float number to avoid dividing by 0.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>t</strong>: A 4D input Tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>m</strong>: A 1D mean Tensor with size matching the last dimension of t.
This is the first output from tf.nn.moments,
or a saved moving average thereof.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>v</strong>: A 1D variance Tensor with size matching the last dimension of t.
This is the second output from tf.nn.moments,
or a saved moving average thereof.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>beta</strong>: A 1D beta Tensor with size matching the last dimension of t.
An offset to be added to the normalized tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>gamma</strong>: A 1D gamma Tensor with size matching the last dimension of t.
If "scale_after_normalization" is true, this tensor will be multiplied
with the normalized tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>result</strong></p></td></tr></table></div><divclass="doc"><p>Batch normalization.</p><p>This op is deprecated. Prefer `tf.nn.batch_normalization`.</p></div></div><divclass="top"><pclass="src"><aname="v:encodeBase64"class="def">encodeBase64</a><ahref="src/TensorFlow-GenOps-Core.html#encodeBase64"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>input</strong>: Strings to be encoded.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>output</strong>: Input strings encoded in base64.</p></td></tr></table></div><divclass="doc"><p>Encode strings into web-safe base64 format.</p><p>Refer to the following article for more information on base64 format:
en.wikipedia.org<em>wiki</em>Base64. Base64 strings may have padding with '=' at the
end so that the encoded has length multiple of 4. See Padding section of the
link above.</p><p>Web-safe means that the encoder uses - and _ instead of + and /.</p></div></div><divclass="top"><pclass="src"><aname="v:stringJoin"class="def">stringJoin</a><ahref="src/TensorFlow-GenOps-Core.html#stringJoin"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>]</td><tdclass="doc"><p><strong>inputs</strong>: A list of string tensors. The tensors must all have the same shape,
or be scalars. Scalars may be mixed in; these will be broadcast to the shape
of non-scalar inputs.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Joins the strings in the given list of string tensors into one tensor;</p><p>with the given separator (default is an empty separator).</p></div></div><divclass="top"><pclass="src"><aname="v:cropAndResizeGradImage"class="def">cropAndResizeGradImage</a><ahref="src/TensorFlow-GenOps-Core.html#cropAndResizeGradImage"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>grads</strong>: A 4-D tensor of shape `[num_boxes, crop_height, crop_width, depth]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>boxes</strong>: A 2-D tensor of shape `[num_boxes, 4]`. The <code>i</code>-th row of the tensor
specifies the coordinates of a box in the `box_ind[i]` image and is specified
in normalized coordinates `[y1, x1, y2, x2]`. A normalized coordinate value of
<code>y</code> is mapped to the image coordinate at `y * (image_height - 1)`, so as the
`[0, 1]` interval of normalized image height is mapped to
`[0, image_height - 1] in image height coordinates. We do allow y1 > y2, in
which case the sampled crop is an up-down flipped version of the original
image. The width dimension is treated similarly. Normalized coordinates
outside the `[0, 1]` range are allowed, in which case we use
<code>extrapolation_value</code> to extrapolate the input image values.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>box_ind</strong>: A 1-D tensor of shape `[num_boxes]` with int32 values in `[0, batch)`.
The value of `box_ind[i]` specifies the image that the <code>i</code>-th box refers to.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>image_size</strong>: A 1-D tensor with value `[batch, image_height, image_width, depth]`
containing the original image size. Both <code>image_height</code> and <code>image_width</code> need
to be positive.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: A 4-D tensor of shape `[batch, image_height, image_width, depth]`.</p></td></tr></table></div><divclass="doc"><p>Computes the gradient of the crop_and_resize op wrt the input image tensor.</p></div></div><divclass="top"><pclass="src"><aname="v:tanh"class="def">tanh</a><ahref="src/TensorFlow-GenOps-Core.html#tanh"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes hyperbolic tangent of <code>x</code> element-wise.</p></div></div><divclass="top"><pclass="src"><aname="v:asString"class="def">asString</a><ahref="src/TensorFlow-GenOps-Core.html#asString"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Converts each entry in the given tensor to strings. Supports many numeric</p><p>types and boolean.</p></div></div><divclass="top"><pclass="src"><aname="v:iFFT2D"class="def">iFFT2D</a><ahref="src/TensorFlow-GenOps-Core.html#iFFT2D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>input</strong>: A complex64 tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> (<ahref=".
dimensions of <code>input</code> are replaced with their inverse 2D Fourier Transform.</p><p><code>compatibility(numpy)
Equivalent to np.ifft2
</code>end_compatibility</p></td></tr></table></div><divclass="doc"><p>Compute the inverse 2-dimensional discrete Fourier Transform over the inner-most</p><p>2 dimensions of <code>input</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseConcat"class="def">sparseConcat</a><ahref="src/TensorFlow-GenOps-Core.html#sparseConcat"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>concat_dim</strong>: Dimension to concatenate along. Must be in range [-rank, rank),
where rank is the number of dimensions in each input <code>SparseTensor</code>.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]</td><tdclass="doc"><p><strong>indices</strong>: 2-D. Indices of each input <code>SparseTensor</code>.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t]</td><tdclass="doc"><p><strong>values</strong>: 1-D. Non-empty values of each <code>SparseTensor</code>.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]</td><tdclass="doc"><p><strong>shapes</strong>: 1-D. Shapes of each <code>SparseTensor</code>.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p>(<strong>output_indices</strong>, <strong>output_values</strong>, <strong>output_shape</strong>)</p><ul><li><strong>output_indices</strong>: 2-D. Indices of the concatenated <code>SparseTensor</code>.</li><li><strong>output_values</strong>: 1-D. Non-empty values of the concatenated <code>SparseTensor</code>.</li><li><strong>output_shape</strong>: 1-D. Shape of the concatenated <code>SparseTensor</code>.</li></ul></td></tr></table></div><divclass="doc"><p>Concatenates a list of <code>SparseTensor</code> along the specified dimension.</p><p>Concatenation is with respect to the dense versions of these sparse tensors.
It is assumed that each input is a <code>SparseTensor</code> whose elements are ordered
along increasing dimension number.</p><p>All inputs' shapes must match, except for the concat dimension. The
<code>indices</code>, <code>values</code>, and <code>shapes</code> lists must have the same length.</p><p>The output shape is identical to the inputs', except along the concat
dimension, where it is the sum of the inputs' sizes along that dimension.</p><p>The output elements will be resorted to preserve the sort order along
increasing dimension number.</p><p>This op runs in `O(M log M)` time, where <code>M</code> is the total number of non-empty
values across all inputs. This is due to the need for an internal sort in
order to concatenate efficiently across an arbitrary dimension.</p><p>For example, if `concat_dim = 1` and the inputs are</p><p>sp_inputs[0]: shape = [2, 3]
[0, 2]: "e"</p><p>then the output will be</p><p>shape = [2, 7]
[0, 2]: "a"
[0, 4]: "d"
[0, 5]: "e"
[1, 0]: "b"
[1, 1]: "c"</p><p>Graphically this is equivalent to doing</p><dl><dt> a</dt><dd>concat [ d e ] = [ a d e ]</dd><dt>b c </dt><dd>[ ] [b c ]</dd></dl></div></div><divclass="top"><pclass="src"><aname="v:shardedFilespec"class="def">shardedFilespec</a><ahref="src/TensorFlow-GenOps-Core.html#shardedFilespec"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>basename</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>num_shards</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>filename</strong></p></td></tr></table></div><divclass="doc"><p>Generate a glob pattern matching all sharded file names.</p></div></div><divclass="top"><pclass="src"><aname="v:transpose"class="def">transpose</a><ahref="src/TensorFlow-GenOps-Core.html#transpose"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tperm, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tperm)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tperm</td><tdclass="doc"><p><strong>perm</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Shuffle dimensions of x according to a permutation.</p><p>The output <code>y</code> has the same rank as <code>x</code>. The shapes of <code>x</code> and <code>y</code> satisfy:
`y.shape[i] == x.shape[perm[i]] for i in [0, 1, ..., rank(x) - 1]`</p></div></div><divclass="top"><pclass="src"><aname="v:reduceJoin"class="def">reduceJoin</a><ahref="src/TensorFlow-GenOps-Core.html#reduceJoin"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>inputs</strong>: The input to be joined. All reduced indices must have non-zero size.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>reduction_indices</strong>: The dimensions to reduce over. Dimensions are reduced in the
order specified. Omitting <code>reduction_indices</code> is equivalent to passing
`[n-1, n-2, ..., 0]`. Negative indices from `-n` to `-1` are supported.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>output</strong>: Has shape equal to that of the input with reduced dimensions removed or
set to `1` depending on <code>keep_dims</code>.</p></td></tr></table></div><divclass="doc"><p>Joins a string Tensor across the given dimensions.</p><p>Computes the string join across dimensions in the given string Tensor of shape
`[d_0, d_1, ..., d_n-1]`. Returns a new Tensor created by joining the input
strings with the given separator (default: empty string). Negative indices are
counted backwards from the end, with `-1` being equivalent to `n - 1`. Passing
an empty <code>reduction_indices</code> joins all strings in linear index order and outputs
a scalar string.</p><p>For example:</p><p>```
# tensor <code>a</code> is [["a", "b"], ["c", "d"]]
```</p></div></div><divclass="top"><pclass="src"><aname="v:stringToHashBucket"class="def">stringToHashBucket</a><ahref="src/TensorFlow-GenOps-Core.html#stringToHashBucket"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_buckets</strong>: The number of buckets.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>string_tensor</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>output</strong>: A Tensor of the same shape as the input <code>string_tensor</code>.</p></td></tr></table></div><divclass="doc"><p>Converts each string in the input Tensor to its hash mod by a number of buckets.</p><p>The hash function is deterministic on the content of the string within the
process.</p><p>Note that the hash function may change from time to time.
This functionality will be deprecated and it's recommended to use
`tf.string_to_hash_bucket_fast()` or `tf.string_to_hash_bucket_strong()`.</p></div></div><divclass="top"><pclass="src"><aname="v:multinomial"class="def">multinomial</a><ahref="src/TensorFlow-GenOps-Core.html#multinomial"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>logits</strong>: 2-D Tensor with shape `[batch_size, num_classes]`. Each slice `[i, :]`
represents the unnormalized log probabilities for all classes.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>num_samples</strong>: 0-D. Number of independent samples to draw for each row slice.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p><strong>output</strong>: 2-D Tensor with shape `[batch_size, num_samples]`. Each slice `[i, :]`
contains the drawn class labels with range `[0, num_classes)`.</p></td></tr></table></div><divclass="doc"><p>Draws samples from a multinomial distribution.</p></div></div><divclass="top"><pclass="src"><aname="v:stringToHashBucketStrong"class="def">stringToHashBucketStrong</a><ahref="src/TensorFlow-GenOps-Core.html#stringToHashBucketStrong"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_buckets</strong>: The number of buckets.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>input</strong>: The strings to assign a hash bucket.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>output</strong>: A Tensor of the same shape as the input <code>string_tensor</code>.</p></td></tr></table></div><divclass="doc"><p>Converts each string in the input Tensor to its hash mod by a number of buckets.</p><p>The hash function is deterministic on the content of the string within the
process. The hash function is a keyed hash function, where attribute <code>key</code>
defines the key of the hash function. <code>key</code> is an array of 2 elements.</p><p>A strong hash is important when inputs may be malicious, e.g. URLs with
additional components. Adversaries could try to make their inputs hash to the
same bucket for a denial-of-service attack or to skew the results. A strong
hash prevents this by making it dificult, if not infeasible, to compute inputs
that hash to the same bucket. This comes at a cost of roughly 4x higher compute
time than `tf.string_to_hash_bucket_fast`.</p></div></div><divclass="top"><pclass="src"><aname="v:scatterNdUpdate"class="def">scatterNdUpdate</a><ahref="src/TensorFlow-GenOps-Core.html#scatterNdUpdate"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ref</strong>: A mutable Tensor. Should be from a Variable node.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>indices</strong>: A Tensor. Must be one of the following types: int32, int64.
A tensor of indices into ref.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>updates</strong>: A Tensor. Must have the same type as ref. A tensor of updated
values to add to ref.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output_ref</strong>: Same as ref. Returned as a convenience for operations that want to
use the updated values after the update is done.</p></td></tr></table></div><divclass="doc"><p>Applies sparse <code>updates</code> to individual values or slices within a given</p><p>variable according to <code>indices</code>.</p><p><code>ref</code> is a <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> with rank <code>P</code> and <code>indices</code> is a <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> of rank <code>Q</code>.</p><p><code>indices</code> must be integer tensor, containing indices into <code>ref</code>.
It must be shape `[d_0, ..., d_{Q-2}, K]` where `0 < K <= P`.</p><p>The innermost dimension of <code>indices</code> (with length <code>K</code>) corresponds to
indices into elements (if `K = P`) or slices (if `K < P`) along the <code>K</code>th
dimension of <code>ref</code>.</p><p><code>updates</code> is <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> of rank `Q-1+P-K` with shape:</p><p>```
print sess.run(update)</p><p>The resulting update to ref would look like this:</p><dl><dt>1, 11, 3, 10, 9, 6, 7, 12</dt><dd></dd></dl><p>See <ahref="#scatter_nd">tf.scatter_nd</a> for more details about how to make updates to
slices.</p></div></div><divclass="top"><pclass="src"><aname="v:fakeQuantWithMinMaxVarsGradient"class="def">fakeQuantWithMinMaxVarsGradient</a><ahref="src/TensorFlow-GenOps-Core.html#fakeQuantWithMinMaxVarsGradient"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>gradients</strong>: Backpropagated gradients above the FakeQuantWithMinMaxVars operation.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>inputs</strong>: Values passed as inputs to the FakeQuantWithMinMaxVars operation.
`gradients * (inputs >= min && inputs <= max)`.</li><li><strong>backprop_wrt_min</strong>: Backpropagated gradients w.r.t. min parameter:
`sum(gradients * (inputs < min))`.</li><li><strong>backprop_wrt_max</strong>: Backpropagated gradients w.r.t. max parameter:
`sum(gradients * (inputs > max))`.</li></ul></td></tr></table></div><divclass="doc"><p>Compute gradients for a FakeQuantWithMinMaxVars operation.</p></div></div><divclass="top"><pclass="src"><aname="v:size"class="def">size</a><ahref="src/TensorFlow-GenOps-Core.html#size"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` out_type)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_type</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Returns the size of a tensor.</p><p>This operation returns an integer representing the number of elements in
```</p></div></div><divclass="top"><pclass="src"><aname="v:scatterDiv"class="def">scatterDiv</a><ahref="src/TensorFlow-GenOps-Core.html#scatterDiv"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ref</strong>: Should be from a <code>Variable</code> node.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>indices</strong>: A tensor of indices into the first dimension of <code>ref</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>updates</strong>: A tensor of values that <code>ref</code> is divided by.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output_ref</strong>: = Same as <code>ref</code>. Returned as a convenience for operations that want
to use the updated values after the update is done.</p></td></tr></table></div><divclass="doc"><p>Divides a variable reference by sparse updates.</p><p>This operation computes</p><p># Scalar indices
ref[indices, ...] /= updates[...]</p><p># Vector indices (for each i)
ref[indices[i], ...] /= updates[i, ...]</p><p># High rank indices (for each i, ..., j)
ref[indices[i, ..., j], ...] /= updates[i, ..., j, ...]</p><p>This operation outputs <code>ref</code> after the update is done.
This makes it easier to chain operations that need to use the reset value.</p><p>Duplicate entries are handled correctly: if multiple <code>indices</code> reference
the same location, their contributions divide.</p><p>Requires `updates.shape = indices.shape + ref.shape[1:]`.</p></div></div><divclass="top"><pclass="src"><aname="v:scatterMul"class="def">scatterMul</a><ahref="src/TensorFlow-GenOps-Core.html#scatterMul"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ref</strong>: Should be from a <code>Variable</code> node.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>indices</strong>: A tensor of indices into the first dimension of <code>ref</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>updates</strong>: A tensor of updated values to multiply to <code>ref</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output_ref</strong>: = Same as <code>ref</code>. Returned as a convenience for operations that want
to use the updated values after the update is done.</p></td></tr></table></div><divclass="doc"><p>Multiplies sparse updates into a variable reference.</p><p>This operation computes</p><p># Scalar indices
ref[indices, ...] *= updates[...]</p><p># Vector indices (for each i)
ref[indices[i], ...] *= updates[i, ...]</p><p># High rank indices (for each i, ..., j)
ref[indices[i, ..., j], ...] *= updates[i, ..., j, ...]</p><p>This operation outputs <code>ref</code> after the update is done.
This makes it easier to chain operations that need to use the reset value.</p><p>Duplicate entries are handled correctly: if multiple <code>indices</code> reference
the same location, their contributions multiply.</p><p>Requires `updates.shape = indices.shape + ref.shape[1:]`.</p></div></div><divclass="top"><pclass="src"><aname="v:copyHost"class="def">copyHost</a><ahref="src/TensorFlow-GenOps-Core.html#copyHost"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Input tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Output tensor, deep-copied from input.</p></td></tr></table></div><divclass="doc"><p>Copy Host Op.</p><p>Performs CPU-to-CPU deep-copying of tensor.</p><p>Unlike the Copy Op, this op has HostMemory constraint on its input or output.</p></div></div><divclass="top"><pclass="src"><aname="v:wholeFileReader"class="def">wholeFileReader</a><ahref="src/TensorFlow-GenOps-Core.html#wholeFileReader"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>)</td><tdclass="doc"><p><strong>reader_handle</strong>: The handle to reference the Reader.</p></td></tr></table></div><divclass="doc"><p>A Reader that outputs the entire contents of a file as a value.</p><p>To use, enqueue filenames in a Queue. The output of ReaderRead will
be a filename (key) and the contents of that file (value).</p></div></div><divclass="top"><pclass="src"><aname="v:takeManySparseFromTensorsMap"class="def">takeManySparseFromTensorsMap</a><ahref="src/TensorFlow-GenOps-Core.html#takeManySparseFromTensorsMap"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sparse_handles</strong>: 1-D, The <code>N</code> serialized <code>SparseTensor</code> objects.
Shape: `[N]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p>(<strong>sparse_indices</strong>, <strong>sparse_values</strong>, <strong>sparse_shape</strong>)</p><ul><li><strong>sparse_indices</strong>: 2-D. The <code>indices</code> of the minibatch <code>SparseTensor</code>.</li><li><strong>sparse_values</strong>: 1-D. The <code>values</code> of the minibatch <code>SparseTensor</code>.</li><li><strong>sparse_shape</strong>: 1-D. The <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> of the minibatch <code>SparseTensor</code>.</li></ul></td></tr></table></div><divclass="doc"><p>Read <code>SparseTensors</code> from a <code>SparseTensorsMap</code> and concatenate them.</p><p>The input <code>sparse_handles</code> must be an <code>int64</code> matrix of shape `[N, 1]` where
<code>N</code> is the minibatch size and the rows correspond to the output handles of
<code>AddSparseToTensorsMap</code> or <code>AddManySparseToTensorsMap</code>. The ranks of the
original <code>SparseTensor</code> objects that went into the given input ops must all
match. When the final <code>SparseTensor</code> is created, it has rank one
(they have been concatenated along a new row dimension on the left).</p><p>The output <code>SparseTensor</code> object's shape values for all dimensions but the
```</p></div></div><divclass="top"><pclass="src"><aname="v:destroyTemporaryVariable"class="def">destroyTemporaryVariable</a><ahref="src/TensorFlow-GenOps-Core.html#destroyTemporaryVariable"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ref</strong>: A reference to the temporary variable tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p><strong>value</strong></p></td></tr></table></div><divclass="doc"><p>Destroys the temporary variable and returns its final value.</p><p>Sets output to the value of the Tensor pointed to by <code>ref</code>, then destroys
the temporary variable called <code>var_name</code>.
All other uses of <code>ref</code> *must* have executed before this op.
This is typically achieved by chaining the ref through each assign op, or by
using control dependencies.</p><p>Outputs the final value of the tensor pointed to by <code>ref</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:assignSub"class="def">assignSub</a><ahref="src/TensorFlow-GenOps-Core.html#assignSub"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ref</strong>: Should be from a <code>Variable</code> node.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>value</strong>: The value to be subtracted to the variable.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output_ref</strong>: = Same as "ref". Returned as a convenience for operations that want
to use the new value after the variable has been updated.</p></td></tr></table></div><divclass="doc"><p>Update <code>ref</code> by subtracting <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code> from it.</p><p>This operation outputs "ref" after the update is done.
This makes it easier to chain operations that need to use the reset value.</p></div></div><divclass="top"><pclass="src"><aname="v:encodeJpeg"class="def">encodeJpeg</a><ahref="src/TensorFlow-GenOps-Core.html#encodeJpeg"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a></td><tdclass="doc"><p><strong>image</strong>: 3-D with shape `[height, width, channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>contents</strong>: 0-D. JPEG-encoded image.</p></td></tr></table></div><divclass="doc"><p>JPEG-encode an image.</p><p><code>image</code> is a 3-D uint8 Tensor of shape `[height, width, channels]`.</p><p>The attr <code>format</code> can be used to override the color format of the encoded
output. Values can be:</p><ul><li>`''`: Use a default format based on the number of channels in the image.</li><li><code>grayscale</code>: Output a grayscale JPEG image. The <code>channels</code> dimension
of <code>image</code> must be 1.</li><li><code>rgb</code>: Output an RGB JPEG image. The <code>channels</code> dimension
of <code>image</code> must be 3.</li></ul><p>If <code>format</code> is not specified or is the empty string, a default format is picked
in function of the number of channels in <code>image</code>:</p><ul><li>1: Output a grayscale image.</li><li>3: Output an RGB image.</li></ul></div></div><divclass="top"><pclass="src"><aname="v:temporaryVariable"class="def">temporaryVariable</a><ahref="src/TensorFlow-GenOps-Core.html#temporaryVariable"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:Shape">Shape</a></td><tdclass="doc"><p><strong>shape</strong>: The shape of the variable tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> dtype)</td><tdclass="doc"><p><strong>ref</strong>: A reference to the variable tensor.</p></td></tr></table></div><divclass="doc"><p>Returns a tensor that may be mutated, but only persists within a single step.</p><p>This is an experimental op for internal use only and it is possible to use this
op in unsafe ways. DO NOT USE unless you fully understand the risks.</p><p>It is the caller's responsibility to ensure that <code>ref</code> is eventually passed to a
matching <code>DestroyTemporaryVariable</code> op after all other uses have completed.</p><p>Outputs a ref to the tensor state so it may be read or modified.</p><p>E.g.
var = state_ops._temporary_variable([1, 2], types.float_)
var_name = var.op.name
var = state_ops.assign(var, [[4.0, 5.0]])
var = state_ops.assign_add(var, [[6.0, 7.0]])
final = state_ops._destroy_temporary_variable(var, var_name=var_name)</p></div></div><divclass="top"><pclass="src"><aname="v:isVariableInitialized"class="def">isVariableInitialized</a><ahref="src/TensorFlow-GenOps-Core.html#isVariableInitialized"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> dtype</td><tdclass="doc"><p><strong>ref</strong>: Should be from a <code>Variable</code> node. May be uninitialized.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a>)</td><tdclass="doc"><p><strong>is_initialized</strong></p></td></tr></table></div><divclass="doc"><p>Checks whether a tensor has been initialized.</p><p>Outputs boolean scalar indicating whether the tensor has been initialized.</p></div></div><divclass="top"><pclass="src"><aname="v:variable"class="def">variable</a><ahref="src/TensorFlow-GenOps-Core.html#variable"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:Shape">Shape</a></td><tdclass="doc"><p><strong>shape</strong>: The shape of the variable tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> dtype)</td><tdclass="doc"><p><strong>ref</strong>: A reference to the variable tensor.</p></td></tr></table></div><divclass="doc"><p>Holds state in the form of a tensor that persists across steps.</p><p>Outputs a ref to the tensor state so it may be read or modified.
TODO(zhifengc/mrry): Adds a pointer to a more detail document
about sharing states in tensorflow.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseSparseMinimum"class="def">sparseSparseMinimum</a><ahref="src/TensorFlow-GenOps-Core.html#sparseSparseMinimum"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>a_indices</strong>: 2-D. `N x R` matrix with the indices of non-empty values in a
SparseTensor, in the canonical lexicographic ordering.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>a_values</strong>: 1-D. <code>N</code> non-empty values corresponding to <code>a_indices</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>a_shape</strong>: 1-D. Shape of the input SparseTensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>b_indices</strong>: counterpart to <code>a_indices</code> for the other operand.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>b_values</strong>: counterpart to <code>a_values</code> for the other operand; must be of the same dtype.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>b_shape</strong>: counterpart to <code>a_shape</code> for the other operand; the two shapes must be equal.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p>(<strong>output_indices</strong>, <strong>output_values</strong>)</p><ul><li><strong>output_indices</strong>: 2-D. The indices of the output SparseTensor.</li><li><strong>output_values</strong>: 1-D. The values of the output SparseTensor.</li></ul></td></tr></table></div><divclass="doc"><p>Returns the element-wise min of two SparseTensors.</p><p>Assumes the two SparseTensors have the same shape, i.e., no broadcasting.</p></div></div><divclass="top"><pclass="src"><aname="v:betainc"class="def">betainc</a><ahref="src/TensorFlow-GenOps-Core.html#betainc"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>a</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>b</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Compute the regularized incomplete beta integral \(I_x(a, b)\).</p><p>The regularized incomplete beta integral is defined as:</p><p>```
beta function.</p></div></div><divclass="top"><pclass="src"><aname="v:assign"class="def">assign</a><ahref="src/TensorFlow-GenOps-Core.html#assign"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ref</strong>: Should be from a <code>Variable</code> node. May be uninitialized.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>value</strong>: The value to be assigned to the variable.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output_ref</strong>: = Same as "ref". Returned as a convenience for operations that want
to use the new value after the variable has been reset.</p></td></tr></table></div><divclass="doc"><p>Update <code>ref</code> by assigning <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code> to it.</p><p>This operation outputs "ref" after the assignment is done.
This makes it easier to chain operations that need to use the reset value.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseSoftmax"class="def">sparseSoftmax</a><ahref="src/TensorFlow-GenOps-Core.html#sparseSoftmax"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sp_indices</strong>: 2-D. `NNZ x R` matrix with the indices of non-empty values in a
SparseTensor, in canonical ordering.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>sp_values</strong>: 1-D. <code>NNZ</code> non-empty values corresponding to <code>sp_indices</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sp_shape</strong>: 1-D. Shape of the input SparseTensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 1-D. The <code>NNZ</code> values for the result <code>SparseTensor</code>.</p></td></tr></table></div><divclass="doc"><p>Applies softmax to a batched N-D <code>SparseTensor</code>.</p><p>The inputs represent an N-D SparseTensor with logical shape `[..., B, C]`
(where `N >= 2`), and with indices sorted in the canonical lexicographic order.</p><p>This op is equivalent to applying the normal `tf.nn.softmax()` to each innermost
logical submatrix with shape `[B, C]`, but with the catch that *the implicitly
zero elements do not participate*. Specifically, the algorithm is equivalent
to the following:</p><ol><li>Applies `tf.nn.softmax()` to a densified view of each innermost submatrix
with shape `[B, C]`, along the size-C dimension;</li><li>Masks out the original implicitly-zero locations;</li><li>Renormalizes the remaining elements.</li></ol><p>Hence, the <code>SparseTensor</code> result has exactly the same non-zero indices and
shape.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseDenseCwiseAdd"class="def">sparseDenseCwiseAdd</a><ahref="src/TensorFlow-GenOps-Core.html#sparseDenseCwiseAdd"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sp_indices</strong>: 2-D. `N x R` matrix with the indices of non-empty values in a
SparseTensor, possibly not in canonical ordering.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>sp_values</strong>: 1-D. <code>N</code> non-empty values corresponding to <code>sp_indices</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sp_shape</strong>: 1-D. Shape of the input SparseTensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>dense</strong>: <code>R</code>-D. The dense Tensor operand.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 1-D. The <code>N</code> values that are operated on.</p></td></tr></table></div><divclass="doc"><p>Adds up a SparseTensor and a dense Tensor, using these special rules:</p><ol><li>Broadcasts the dense side to have the same shape as the sparse side, if
eligible;</li><li>Then, only the dense values pointed to by the indices of the SparseTensor
participate in the cwise addition.</li></ol><p>By these rules, the result is a logical SparseTensor with exactly the same
indices and shape, but possibly with different non-zero values. The output of
this Op is the resultant non-zero values.</p></div></div><divclass="top"><pclass="src"><aname="v:logicalNot"class="def">logicalNot</a><ahref="src/TensorFlow-GenOps-Core.html#logicalNot"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Returns the truth value of NOT x element-wise.</p></div></div><divclass="top"><pclass="src"><aname="v:queueSize"class="def">queueSize</a><ahref="src/TensorFlow-GenOps-Core.html#queueSize"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a queue.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>)</td><tdclass="doc"><p><strong>size</strong>: The number of elements in the given queue.</p></td></tr></table></div><divclass="doc"><p>Computes the number of elements in the given queue.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseApplyAdagrad"class="def">sparseApplyAdagrad</a><ahref="src/TensorFlow-GenOps-Core.html#sparseApplyAdagrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>accum</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc
accum += grad * grad
var -= lr * grad * (1 / sqrt(accum))</p></div></div><divclass="top"><pclass="src"><aname="v:getSessionHandle"class="def">getSessionHandle</a><ahref="src/TensorFlow-GenOps-Core.html#getSessionHandle"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>value</strong>: The tensor to be stored.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle for the tensor stored in the session state.</p></td></tr></table></div><divclass="doc"><p>Store the input tensor in the state of the current session.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseDenseCwiseMul"class="def">sparseDenseCwiseMul</a><ahref="src/TensorFlow-GenOps-Core.html#sparseDenseCwiseMul"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sp_indices</strong>: 2-D. `N x R` matrix with the indices of non-empty values in a
SparseTensor, possibly not in canonical ordering.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>sp_values</strong>: 1-D. <code>N</code> non-empty values corresponding to <code>sp_indices</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sp_shape</strong>: 1-D. Shape of the input SparseTensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>dense</strong>: <code>R</code>-D. The dense Tensor operand.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 1-D. The <code>N</code> values that are operated on.</p></td></tr></table></div><divclass="doc"><p>Component-wise multiplies a SparseTensor by a dense Tensor.</p><p>The output locations corresponding to the implicitly zero elements in the sparse
tensor will be zero (i.e., will not take up storage space), regardless of the
contents of the dense tensor (even if it's +/-INF and that INF*0 == NaN).</p><ul><li>Limitation*: this Op only broadcasts the dense side to the sparse side, but not
the other direction.</li></ul></div></div><divclass="top"><pclass="src"><aname="v:sparseTensorDenseAdd"class="def">sparseTensorDenseAdd</a><ahref="src/TensorFlow-GenOps-Core.html#sparseTensorDenseAdd"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 tindices</td><tdclass="doc"><p><strong>a_indices</strong>: 2-D. The <code>indices</code> of the <code>SparseTensor</code>, with shape `[nnz, ndims]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>a_values</strong>: 1-D. The <code>values</code> of the <code>SparseTensor</code>, with shape `[nnz]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 tindices</td><tdclass="doc"><p><strong>a_shape</strong>: 1-D. The <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> of the <code>SparseTensor</code>, with shape `[ndims]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>b</strong>: <code>ndims</code>-D Tensor. With shape <code>a_shape</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Adds up a <code>SparseTensor</code> and a dense <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code>, producing a dense <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code>.</p><p>This Op does not require <code>a_indices</code> be sorted in standard lexicographic order.</p></div></div><divclass="top"><pclass="src"><aname="v:getSessionTensor"class="def">getSessionTensor</a><ahref="src/TensorFlow-GenOps-Core.html#getSessionTensor"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle for a tensor stored in the session state.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value
SparseTensor, possibly not in canonical ordering.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>input_values</strong>: 1-D. <code>N</code> non-empty values corresponding to <code>input_indices</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>input_shape</strong>: 1-D. Shape of the input SparseTensor.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p>(<strong>output_indices</strong>, <strong>output_values</strong>)</p><ul><li><strong>output_indices</strong>: 2-D. `N x R` matrix with the same indices as input_indices, but
in canonical row-major ordering.</li><li><strong>output_values</strong>: 1-D. <code>N</code> non-empty values corresponding to <code>output_indices</code>.</li></ul></td></tr></table></div><divclass="doc"><p>Reorders a SparseTensor into the canonical, row-major ordering.</p><p>Note that by convention, all sparse ops preserve the canonical ordering along
increasing dimension number. The only time ordering can be violated is during
manual manipulation of the indices and values vectors to add entries.</p><p>Reordering does not affect the shape of the SparseTensor.</p><p>If the tensor has rank <code>R</code> and <code>N</code> non-empty values, <code>input_indices</code> has
shape `[N, R]`, input_values has length <code>N</code>, and input_shape has length <code>R</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseSplit"class="def">sparseSplit</a><ahref="src/TensorFlow-GenOps-Core.html#sparseSplit"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_split</strong>: The number of ways to split.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>split_dim</strong>: 0-D. The dimension along which to split. Must be in the range
`[0, rank(shape))`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>indices</strong>: 2-D tensor represents the indices of the sparse tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>values</strong>: 1-D tensor represents the values of the sparse tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>shape</strong>: 1-D. tensor represents the shape of the sparse tensor.
output indices: A list of 1-D tensors represents the indices of the output
sparse tensors.</p></td></tr><tr><tdclass="src">-> ([<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>], [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t], [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>])</td><tdclass="doc"><p>(<strong>output_indices</strong>, <strong>output_values</strong>, <strong>output_shape</strong>)</p><ul><li><strong>output_indices</strong></li><li><strong>output_values</strong>: A list of 1-D tensors represents the values of the output sparse
tensors.</li><li><strong>output_shape</strong>: A list of 1-D tensors represents the shape of the output sparse
tensors.</li></ul></td></tr></table></div><divclass="doc"><p>Split a <code>SparseTensor</code> into <code>num_split</code> tensors along one dimension.</p><p>If the `shape[split_dim]` is not an integer multiple of <code>num_split</code>. Slices
`[0 : shape[split_dim] % num_split]` gets one extra dimension.
For example, if `split_dim = 1` and `num_split = 2` and the input is</p><p>input_tensor = shape = [2, 7]
[ a d e ]
[b c ]</p><p>Graphically the output tensors are:</p><p>output_tensor[0] = shape = [2, 4]
[ a ]
[b c ]</p><p>output_tensor[1] = shape = [2, 3]
[ d e ]
[ ]</p></div></div><divclass="top"><pclass="src"><aname="v:pad"class="def">pad</a><ahref="src/TensorFlow-GenOps-Core.html#pad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tpaddings, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tpaddings)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tpaddings</td><tdclass="doc"><p><strong>paddings</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Pads a tensor with zeros.</p><p>This operation pads a <code>input</code> with zeros according to the <code>paddings</code> you
specify. <code>paddings</code> is an integer tensor with shape `[Dn, 2]`, where n is the
rank of <code>input</code>. For each dimension D of <code>input</code>, `paddings[D, 0]` indicates
how many zeros to add before the contents of <code>input</code> in that dimension, and
`paddings[D, 1]` indicates how many zeros to add after the contents of <code>input</code>
in that dimension.</p><p>The padded size of each dimension D of the output is:</p><p>`paddings(D, 0) + input.dim_size(D) + paddings(D, 1)`</p><p>For example:</p><p>```prettyprint
index where `sparse_values[i]` will be placed.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>output_shape</strong>: 1-D. Shape of the dense output tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>sparse_values</strong>: 1-D. Values corresponding to each row of <code>sparse_indices</code>,
or a scalar value to be used for all sparse indices.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>default_value</strong>: Scalar value to set for indices not specified in
<code>sparse_indices</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>dense</strong>: Dense output tensor of shape <code>output_shape</code>.</p></td></tr></table></div><divclass="doc"><p>Converts a sparse representation into a dense tensor.</p><p>Builds an array <code>dense</code> with shape <code>output_shape</code> such that</p><p>```prettyprint
# If sparse_indices is scalar
dense[i] = (i == sparse_indices ? sparse_values : default_value)</p><p># If sparse_indices is a vector, then for each i
dense[sparse_indices[i]] = sparse_values[i]</p><p># If sparse_indices is an n by d matrix, then for each i in [0, n)
```</p><p>All other values in <code>dense</code> are set to <code>default_value</code>. If <code>sparse_values</code> is a
scalar, all sparse indices are set to this single value.</p><p>Indices should be sorted in lexicographic order, and indices must not
contain any repeats. If <code>validate_indices</code> is true, these properties
are checked during execution.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseTensorDenseMatMul"class="def">sparseTensorDenseMatMul</a><ahref="src/TensorFlow-GenOps-Core.html#sparseTensorDenseMatMul"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>a_indices</strong>: 2-D. The <code>indices</code> of the <code>SparseTensor</code>, size `[nnz, 2]` Matrix.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>a_values</strong>: 1-D. The <code>values</code> of the <code>SparseTensor</code>, size `[nnz]` Vector.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>a_shape</strong>: 1-D. The <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> of the <code>SparseTensor</code>, size `[2]` Vector.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>b</strong>: 2-D. A dense Matrix.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>product</strong></p></td></tr></table></div><divclass="doc"><p>Multiply SparseTensor (of rank 2) <ahref="A.html">A</a> by dense matrix <ahref="B.html">B</a>.</p><p>No validity checking is performed on the indices of A. However, the following
input format is recommended for optimal behavior:</p><p>if adjoint_a == false:
A should be sorted in lexicographically increasing order. Use SparseReorder
if you're not sure.
if adjoint_a == true:
A should be sorted in order of increasing dimension 1 (i.e., "column major"
order instead of "row major" order).</p></div></div><divclass="top"><pclass="src"><aname="v:mirrorPadGrad"class="def">mirrorPadGrad</a><ahref="src/TensorFlow-GenOps-Core.html#mirrorPadGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tpaddings, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tpaddings)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: The input tensor to be folded.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tpaddings</td><tdclass="doc"><p><strong>paddings</strong>: A two-column matrix specifying the padding sizes. The number of
rows must be the same as the rank of <code>input</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The folded tensor.</p></td></tr></table></div><divclass="doc"><p>Gradient op for <code>MirrorPad</code> op. This op folds a mirror-padded tensor.</p><p>This operation folds the padded areas of <code>input</code> by <code>MirrorPad</code> according to the
<code>paddings</code> you specify. <code>paddings</code> must be the same as <code>paddings</code> argument
given to the corresponding <code>MirrorPad</code> op.</p><p>The folded size of each dimension D of the output is:</p><p>`input.dim_size(D) - paddings(D, 0) - paddings(D, 1)`</p><p>For example:</p><p>```prettyprint
```</p></div></div><divclass="top"><pclass="src"><aname="v:randomShuffle"class="def">randomShuffle</a><ahref="src/TensorFlow-GenOps-Core.html#randomShuffle"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>value</strong>: The tensor to be shuffled.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p><strong>output</strong>: A tensor of same shape and type as <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>, shuffled along its first
dimension.</p></td></tr></table></div><divclass="doc"><p>Randomly shuffles a tensor along its first dimension.</p><p>The tensor is shuffled along dimension 0, such that each `value[j]` is mapped
to one and only one `output[i]`. For example, a mapping that might occur for a
3x2 tensor is:</p><p>```prettyprint
[[1, 2], [[5, 6],
[3, 4], ==> [1, 2],
[5, 6]] [3, 4]]
```</p></div></div><divclass="top"><pclass="src"><aname="v:select"class="def">select</a><ahref="src/TensorFlow-GenOps-Core.html#select"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>condition</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>t</strong>: = A <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> which may have the same shape as <code>condition</code>.
If <code>condition</code> is rank 1, <code>t</code> may have higher rank,
but its first dimension must match the size of <code>condition</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>e</strong>: = A <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> with the same type and shape as <code>t</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: = A <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> with the same type and shape as <code>t</code> and <code>e</code>.</p></td></tr></table></div><divclass="doc"><p>Selects elements from <code>t</code> or <code>e</code>, depending on <code>condition</code>.</p><p>The <code>t</code>, and <code>e</code> tensors must all have the same shape, and the
output will also have that shape.</p><p>The <code>condition</code> tensor must be a scalar if <code>t</code> and <code>e</code> are scalars.
If <code>t</code> and <code>e</code> are vectors or higher rank, then <code>condition</code> must be either a
scalar, a vector with size matching the first dimension of <code>t</code>, or must have
the same shape as <code>t</code>.</p><p>The <code>condition</code> tensor acts as a mask that chooses, based on the value at each
element, whether the corresponding element / row in the output should be
taken from <code>t</code> (if true) or <code>e</code> (if false).</p><p>If <code>condition</code> is a vector and <code>t</code> and <code>e</code> are higher rank matrices, then
it chooses which row (outer dimension) to copy from <code>t</code> and <code>e</code>.
If <code>condition</code> has the same shape as <code>t</code> and <code>e</code>, then it chooses which
element to copy from <code>t</code> and <code>e</code>.</p><p>For example:</p><p>```prettyprint
# <code>condition</code> tensor is [[True, False]
# [False, True]]
# <code>t</code> is [[1, 2],
# [3, 4]]
# <code>e</code> is [[5, 6],
# [7, 8]]
select(condition, t, e) ==> [[1, 6],
[7, 4]]</p><p># <code>condition</code> tensor is [True, False]
# <code>t</code> is [[1, 2],
# [3, 4]]
# <code>e</code> is [[5, 6],
# [7, 8]]
select(condition, t, e) ==> [[1, 2],
[7, 8]]</p><p>```</p></div></div><divclass="top"><pclass="src"><aname="v:sparseAddGrad"class="def">sparseAddGrad</a><ahref="src/TensorFlow-GenOps-Core.html#sparseAddGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>backprop_val_grad</strong>: 1-D with shape `[nnz(sum)]`. The gradient with respect to
the non-empty values of the sum.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>a_indices</strong>: 2-D. The <code>indices</code> of the <code>SparseTensor</code> A, size `[nnz(A), ndims]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>b_indices</strong>: 2-D. The <code>indices</code> of the <code>SparseTensor</code> B, size `[nnz(B), ndims]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sum_indices</strong>: 2-D. The <code>indices</code> of the sum <code>SparseTensor</code>, size
`[nnz(sum), ndims]`.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p>(<strong>a_val_grad</strong>, <strong>b_val_grad</strong>)</p><ul><li><strong>a_val_grad</strong>: 1-D with shape `[nnz(A)]`. The gradient with respect to the
non-empty values of A.</li><li><strong>b_val_grad</strong>: 1-D with shape `[nnz(B)]`. The gradient with respect to the
non-empty values of B.</li></ul></td></tr></table></div><divclass="doc"><p>The gradient operator for the SparseAdd op.</p><p>The SparseAdd op calculates A + B, where A, B, and the sum are all represented
as <code>SparseTensor</code> objects. This op takes in the upstream gradient w.r.t.
non-empty values of the sum, and outputs the gradients w.r.t. the non-empty
values of A and B.</p></div></div><divclass="top"><pclass="src"><aname="v:sdcaFprint"class="def">sdcaFprint</a><ahref="src/TensorFlow-GenOps-Core.html#sdcaFprint"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>input</strong>: vector of strings to compute fingerprints on.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>output</strong>: a (N,2) shaped matrix where N is the number of elements in the input
vector. Each row contains the low and high parts of the fingerprint.</p></td></tr></table></div><divclass="doc"><p>Computes fingerprints of the input strings.</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArrayUnpack"class="def">tensorArrayUnpack</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArrayUnpack"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>value</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>flow_out</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:quantizedAvgPool"class="def">quantizedAvgPool</a><ahref="src/TensorFlow-GenOps-Core.html#quantizedAvgPool"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: 4-D with shape `[batch, height, width, channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>min_input</strong>: The float value that the lowest quantized input value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>max_input</strong>: The float value that the highest quantized input value represents.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>output</strong>, <strong>min_output</strong>, <strong>max_output</strong>)</p><ul><li><strong>output</strong></li><li><strong>min_output</strong>: The float value that the lowest quantized output value represents.</li><li><strong>max_output</strong>: The float value that the highest quantized ou
interpreted as `[height, width, channels]`. The other dimensions only
represent a collection of images, such as `[batch, height, width, channels].`</p><p>Contrast is adjusted independently for each channel of each image.</p><p>For each channel, the Op first computes the mean of the image pixels in the
channel and then adjusts each component of each pixel to
`(x - mean) * contrast_factor + mean`.</p></div></div><divclass="top"><pclass="src"><aname="v:resourceGather"class="def">resourceGather</a><ahref="src/TensorFlow-GenOps-Core.html#resourceGather"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ResourceHandle">ResourceHandle</a> dtype</td><tdclass="doc"><p><strong>resource</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>indices</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype)</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Gather slices from the variable pointed to by <code>resource</code> according to <code>indices</code>.</p><p><code>indices</code> must be an integer tensor of any dimension (usually 0-D or 1-D).
Produces an output tensor with shape `indices.shape + params.shape[1:]` where:</p><p>```python
# Scalar indices
output[:, ..., :] = params[indices, :, ... :]</p><p># Vector indices
```</p></div></div><divclass="top"><pclass="src"><aname="v:mergeSummary"class="def">mergeSummary</a><ahref="src/TensorFlow-GenOps-Core.html#mergeSummary"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>]</td><tdclass="doc"><p><strong>inputs</strong>: Can be of any shape. Each must contain serialized <code>Summary</code> protocol
buffers.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>summary</strong>: Scalar. Serialized <code>Summary</code> protocol buffer.</p></td></tr></table></div><divclass="doc"><p>Merges summaries.</p><p>This op creates a
protocol buffer that contains the union of all the values in the input
summaries.</p><p>When the Op is run, it reports an <code>InvalidArgument</code> error if multiple values
in the summaries to merge use the same tag.</p></div></div><divclass="top"><pclass="src"><aname="v:serializeSparse"class="def">serializeSparse</a><ahref="src/TensorFlow-GenOps-Core.html#serializeSparse"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sparse_indices</strong>: 2-D. The <code>indices</code> of the <code>SparseTensor</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>sparse_values</strong>: 1-D. The <code>values</code> of the <code>SparseTensor</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sparse_shape</strong>: 1-D. The <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> of the <code>SparseTensor</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>serialized_sparse</strong></p></td></tr></table></div><divclass="doc"><p>Serialize a <code>SparseTensor</code> into a string 3-vector (1-D <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code>) object.</p></div></div><divclass="top"><pclass="src"><aname="v:negTrain"class="def">negTrain</a><ahref="src/TensorFlow-GenOps-Core.html#negTrain"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_negative_samples</strong>: Number of negative samples per example.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>w_in</strong>: input word embedding.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>w_out</strong>: output word embedding.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>examples</strong>: A vector of word ids.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>labels</strong>: A vector of word ids.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>lr</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Training via negative sampling.</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArrayCloseV2"class="def">tensorArrayCloseV2</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArrayCloseV2"class="link">Sou
candidates in a batch are unique. This requires some approximation to
estimate the post-rejection sampling probabilities.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>true_classes</strong>: A batch_size * num_true matrix, in which each row contains the
IDs of the num_true target_classes in the corresponding original label.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>sampled_candidates</strong>, <strong>true_expected_count</strong>, <strong>sampled_expected_count</strong>)</p><ul><li><strong>sampled_candidates</strong>: A vector of length num_sampled, in which each element is
the ID of a sampled candidate.</li><li><strong>true_expected_count</strong>: A batch_size * num_true matrix, representing
the number of times each candidate is expected to occur in a batch
of sampled candidates. If unique=true, then this is a probability.</li><li><strong>sampled_expected_count</strong>: A vector of length num_sampled, for each sampled
candidate representing the number of times the candidate is expected
to occur in a batch of sampled candidates. If unique=true, then this is a
probability.</li></ul></td></tr></table></div><divclass="doc"><p>Generates labels for candidate sampling with a learned unigram distribution.</p><p>See explanations of candidate sampling and the data formats at
go/candidate-sampling.</p><p>For each batch, this op picks a single set of sampled candidate labels.</p><p>The advantages of sampling candidates per-batch are simplicity and the
possibility of efficient dense matrix multiplication. The disadvantage is that
the sampled candidates must be chosen independently of the context and of the
true labels.</p></div></div><divclass="top"><pclass="src"><aname="v:stringToNumber"class="def">stringToNumber</a><ahref="src/TensorFlow-GenOps-Core.html#stringToNumber"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` out_type)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>string_tensor</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_type</td><tdclass="doc"><p><strong>output</strong>: A Tensor of the same shape as the input <code>string_tensor</code>.</p></td></tr></table></div><divclass="doc"><p>Converts each string in the input Tensor to the specified numeric type.</p><p>(Note that int32 overflow results in an error while float overflow
results in a rounded value.)</p></div></div><divclass="top"><pclass="src"><aname="v:cTCBeamSearchDecoder"class="def">cTCBeamSearchDecoder</a><ahref="src/TensorFlow-GenOps-Core.html#cTCBeamSearchDecoder"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>beam_width</strong>: A scalar >= 0 (beam search beam width).</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>top_paths</strong>: A scalar >= 0, <= beam_width (controls output size).</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>inputs</strong>: 3-D, shape: `(max_time x batch_size x num_classes)`, the logits.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>sequence_length</strong>: A vector containing sequence lengths, size `(batch)`.</p></td></tr><tr><tdclass="src">-> ([<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>], [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>], [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>], <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>decoded_indices</strong>, <strong>decoded_values</strong>, <strong>decoded_shape</strong>, <strong>log_probability</strong>)</p><ul><li><strong>decoded_indices</strong>: A list (length: top_paths) of indices matrices. Matrix j,
size `(total_decoded_outputs[j] x 2)`, has indices of a
`SparseTensor<ahref="int64,">2</a>`. The rows store: [batch, time].</li><li><strong>decoded_values</strong>: A list (length: top_paths) of values vectors. Vector j,
size `(length total_decoded_outputs[j])`, has the values of a
`SparseTensor<ahref="int64,">2</a>`. The vector stores the decoded classes for beam j.</li><li><strong>decoded_shape</strong>: A list (length: top_paths) of shape vector. Vector j,
size `(2)`, stores the shape of the decoded `SparseTensor[j]`.
Its values are: `[batch_size, max_decoded_length[j]]`.</li><li><strong>log_probability</strong>: A matrix, shaped: `(batch_size x top_paths)`. The
sequence log-probabilities.</li></ul></td></tr></table></div><divclass="doc"><p>Performs beam search decoding on the logits given in input.</p><p>A note about the attribute merge_repeated: For the beam search decoder,
this means that if consecutive entries in a beam are the same, only
the first of these is emitted. That is, when the top path is "A B B B B",
"A B" is returned if merge_repeated = True but "A B B B B" is
returned if merge_repeated = False.</p></div></div><divclass="top"><pclass="src"><aname="v:parseTensor"class="def">parseTensor</a><ahref="src/TensorFlow-GenOps-Core.html#parseTensor"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> out_type</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>serialized</strong>: A scalar string containing a serialized TensorProto proto.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_type</td><tdclass="doc"><p><strong>output</strong>: A Tensor of type <code>out_type</code>.</p></td></tr></table></div><divclass="doc"><p>Transforms a serialized tensorflow.TensorProto proto into a Tensor.</p></div></div><divclass="top"><pclass="src"><aname="v:imageSummary"class="def">imageSummary</a><ahref="src/TensorFlow-GenOps-Core.html#imageSummary"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>tag</strong>: Scalar. Used to build the <code>tag</code> attribute of the summary values.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>tensor</strong>: 4-D of shape `[batch_size, height, width, channels]` where
<code>channels</code> is 1, 3, or 4.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>summary</strong>: Scalar. Serialized <code>Summary</code> protocol buffer.</p></td></tr></table></div><divclass="doc"><p>Outputs a <code>Summary</code> protocol buffer with images.</p><p>The summary has up to <code>max_images</code> summary values containing images. The
images are built from <code>tensor</code> which must be 4-D with shape `[batch_size,
height, width, channels]` and where <code>channels</code> can be:</p><ul><li>1: <code>tensor</code> is interpreted as Grayscale.</li><li>3: <code>tensor</code> is interpreted as RGB.</li><li>4: <code>tensor</code> is interpreted as RGBA.</li></ul><p>The images have the same number of channels as the input tensor. For float
input, the values are normalized one image at a time to fit in the range
`[0, 255]`. <code>uint8</code> values are unchanged. The op uses two different
normalization algorithms:</p><ul><li>If the input values are all positive, they are rescaled so the largest one
is 255.</li><li>If any input value is negative, the values are shifted so input value 0.0
is at 127. They are then rescaled so that either the smallest value is 0,
or the largest one is 255.</li></ul><p>The <code>tag</code> argument is a scalar <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> of type <code>string</code>. It is used to
build the <code>tag</code> of the summary values:</p><ul><li>If <code>max_images</code> is 1, the summary value tag is '*tag*/image'.</li><li>If <code>max_images</code> is greater than 1, the summary value tags are
generated sequentially as '*tag*/image/0', '*tag*/image/1', etc.</li></ul><p>The <code>bad_color</code> argument is the color to use in the generated images for
non-finite input values. It is a <code>unit8</code> 1-D tensor of length <code>channels</code>.
Each element must be in the range `[0, 255]` (It represents the value of a
pixel in the output image). Non-finite values in the input tensor are
replaced by this tensor in the output image. The default value is the color
red.</p></div></div><divclass="top"><pclass="src"><aname="v:truncateDiv"class="def">truncateDiv</a><ahref="src/TensorFlow-GenOps-Core.html#truncateDiv"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns x / y element-wise for integer types.</p><p>Truncation designates that negative numbers will round fractional quantities
toward zero. I.e. -7 / 5 = 1. This matches C semantics but it is different
than Python semantics. See <code>FloorDiv</code> for a division function that matches
Python Semantics.</p><ul><li>NOTE*: <code>TruncateDiv</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:cholesky"class="def">cholesky</a><ahref="src/TensorFlow-GenOps-Core.html#cholesky"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Shape is `[..., M, M]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Shape is `[..., M, M]`.</p></td></tr></table></div><divclass="doc"><p>Computes the Cholesky decomposition of one or more square matrices.</p><p>The input is a tensor of shape `[..., M, M]` whose inner-most 2 dimensions
form square matrices, with the same constraints as the single matrix Cholesky
decomposition above. The output is a tensor of the same shape as the input
containing the Cholesky decompositions for all input submatrices `[..., :, :]`.</p></div></div><divclass="top"><pclass="src"><aname="v:batchMatrixSolveLs"class="def">batchMatrixSolveLs</a><ahref="src/TensorFlow-GenOps-Core.html#batchMatrixSolveLs"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>matrix</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>rhs</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a></td><tdclass="doc"><p><strong>l2_regularizer</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:lookupTableExport"class="def">lookupTableExport</a><ahref="src/TensorFlow-GenOps-Core.html#lookupTableExport"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tkeys, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tvalues)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>table_handle</strong>: Handle to the table.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> tkeys, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> tvalues)</td><tdclass="doc"><p>(<strong>keys</strong>, <strong>values</strong>)</p><ul><li><strong>keys</strong>: Vector of all keys present in the table.</li><li><strong>values</strong>: Tensor of all values in the table. Indexed in parallel with <code>keys</code>.</li></ul></td></tr></table></div><divclass="doc"><p>Outputs all keys and values in the table.</p></div></div><divclass="top"><pclass="src"><aname="v:batchSvd"class="def">batchSvd</a><ahref="src/TensorFlow-GenOps-Core.html#batchSvd"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</str
new size for the images.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>resized_images</strong>: 4-D with shape
`[batch, new_height, new_width, channels]`.</p></td></tr></table></div><divclass="doc"><p>Resize <code>images</code> to <code><ahref="TensorFlow-GenOps-Core.html#v:size">size</a></code> using bicubic interpolation.</p><p>Input images can be of different types but output images are always float.</p></div></div><divclass="top"><pclass="src"><aname="v:hSVToRGB"class="def">hSVToRGB</a><ahref="src/TensorFlow-GenOps-Core.html#hSVToRGB"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>images</strong>: 1-D or higher rank. HSV data to convert. Last dimension must be size 3.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: <code>images</code> converted to RGB.</p></td></tr></table></div><divclass="doc"><p>Convert one or more images from HSV to RGB.</p><p>Outputs a tensor of the same shape as the <code>images</code> tensor, containing the RGB
value of the pixels. The output is only well defined if the value in <code>images</code>
are in `[0,1]`.</p><p>See <code>rgb_to_hsv</code> for a description of the HSV encoding.</p></div></div><divclass="top"><pclass="src"><aname="v:avgPool3D"class="def">avgPool3D</a><ahref="src/TensorFlow-GenOps-Core.html#avgPool3D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Shape `[batch, depth, rows, cols, channels]` tensor to pool over.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The average pooled output tensor.</p></td></tr></table></div><divclass="doc"><p>Performs 3D average pooling on the input.</p></div></div><divclass="top"><pclass="src"><aname="v:stackClose"class="def">stackClose</a><ahref="src/TensorFlow-GenOps-Core.html#stackClose"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a stack.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Delete the stack from its resource container.</p></div></div><divclass="top"><pclass="src"><aname="v:assignVariableOp"class="def">assignVariableOp</a><ahref="src/TensorFlow-GenOps-Core.html#assignVariableOp"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ResourceHandle">ResourceHandle</a> dtype</td><tdclass="doc"><p><strong>resource</strong>: handle to the resource in which to store the variable.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 dtype</td><tdclass="doc"><p><strong>value</strong>: the value to set the new tensor to use.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Assigns a new value to a variable.</p><p>Any ReadVariableOp with a control dependency on this op is guaranteed to return
this value or a subsequent newer value of the variable.</p></div></div><divclass="top"><pclass="src"><aname="v:lRN"class="def">lRN</a><ahref="src/TensorFlow-GenOps-Core.html#lRN"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: 4-D.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Local Response Normalization.</p><p>The 4-D <code>input</code> tensor is treated as a 3-D array of 1-D vectors (along the last
dimension), and each vector is normalized independently. Within a given vector,
each component is divided by the weighted, squared sum of inputs within
<code>depth_radius</code>. In detail,</p><p>sqr_sum[a, b, c, d] =
sum(input[a, b, c, d - depth_radius : d + depth_radius + 1] ** 2)
output = input / (bias + alpha * sqr_sum) ** beta</p><p>For details, see <ahref="http://papers.nips.cc/paper/4824-imagenet-classification-with-deep-convolutional-neural-networks">Krizhevsky et al., ImageNet classification with deep
convolutional neural networks (NIPS 2012)</a>.</p></div></div><divclass="top"><pclass="src"><aname="v:zeta"class="def">zeta</a><ahref="src/TensorFlow-GenOps-Core.html#zeta"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>q</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Compute the Hurwitz zeta function \(zeta(x, q)\).</p><p>The Hurwitz zeta function is defined as:</p><p>```
zeta(x, q) = sum_{n=0}^{infty} (q + n)^{-x}
```</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArrayGradV2"class="def">tensorArrayGradV2</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArrayGradV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to the forward TensorArray.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong>: A float scalar that enforces proper chaining of operations.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>)</td><tdclass="doc"><p><strong>grad_handle</strong></p></td></tr></table></div><divclass="doc"><p>Creates a TensorArray for storing the gradients of values in the given handle.</p><p>If the given TensorArray gradient already exists, returns a reference to it.</p><p>Locks the size of the original TensorArray by disabling its dynamic size flag.</p><ul><li>*A note about the input flow_in:**</li></ul><p>The handle flow_in forces the execution of the gradient lookup to occur
only after certain other operations have occurred. For example, when
the forward TensorArray is dynamically sized, writes to this TensorArray
may resize the object. The gradient TensorArray is statically sized based
on the size of the forward TensorArray when this operation executes.
Furthermore, the size of the forward TensorArray is frozen by this call.
As a result, the flow is used to ensure that the call to generate the gradient
TensorArray only happens after all writes are executed.</p><p>In the case of dynamically sized TensorArrays, gradient computation should
only be performed on read operations that have themselves been chained via
flow to occur only after all writes have executed. That way the final size
of the forward TensorArray is known when this operation is called.</p><ul><li>*A note about the source attribute:**</li></ul><p>TensorArray gradient calls use an accumulator TensorArray object. If
multiple gradients are calculated and run in the same session, the multiple
gradient nodes may accidentally flow throuth the same accumulator TensorArray.
This double counts and generally breaks the TensorArray gradient flow.</p><p>The solution is to identify which gradient call this particular
TensorArray gradient is being called in. This is performed by identifying
a unique string (e.g. "gradients", "gradients_1", ...) from the input
gradient Tensor's name. This string is used as a suffix when creating
the TensorArray gradient object here (the attribute <code>source</code>).</p><p>The attribute <code>source</code> is added as a suffix to the forward TensorArray's
name when performing the creation / lookup, so that each separate gradient
calculation gets its own TensorArray accumulator.</p></div></div><divclass="top"><pclass="src"><aname="v:cast"class="def">cast</a><ahref="src/TensorFlow-GenOps-Core.html#cast"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> srcT, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dstT)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 srcT</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dstT</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Cast x of type SrcT to y of DstT.</p></div></div><divclass="top"><pclass="src"><aname="v:erf"class="def">erf</a><ahref="src/TensorFlow-GenOps-Core.html#erf"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Computes the Gauss error function of <code>x</code> element-wise.</p></div></div><divclass="top"><pclass="src"><aname="v:batchMatrixTriangularSolve"class="def">batchMatrixTriangularSolve</a><ahref="src/TensorFlow-GenOps-Core.html#batchMatrixTriangularSolve"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>matrix</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>rhs</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:resourceScatterAdd"class="def">resourceScatterAdd</a><ahref="src/TensorFlow-GenOps-Core.html#resourceScatterAdd"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../b
ref[indices, ...] += updates[...]</p><p># Vector indices (for each i)
ref[indices[i], ...] += updates[i, ...]</p><p># High rank indices (for each i, ..., j)
ref[indices[i, ..., j], ...] += updates[i, ..., j, ...]</p><p>Duplicate entries are handled correctly: if multiple <code>indices</code> reference
the same location, their contributions add.</p><p>Requires `updates.shape = indices.shape + ref.shape[1:]`.</p><p><ahref="div">style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;"</a>
form matrices of size `[M, N]`. Let <code>P</code> be the minimum of <code>M</code> and <code>N</code>.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p>(<strong>s</strong>, <strong>u</strong>, <strong>v</strong>)</p><ul><li><strong>s</strong>: Singular values. Shape is `[..., P]`.</li><li><strong>u</strong>: Left singular vectors. If <code>full_matrices</code> is <code><ahref="../base-4.8.2.0/Data-Bool.html#v:False">False</a></code> then shape is
`[..., M, M]`; if <code>full_matrices</code> is <code><ahref="../base-4.8.2.0/Data-Bool.html#v:True">True</a></code> then shape is
`[..., M, P]`. Undefined if <code>compute_uv</code> is <code><ahref="../base-4.8.2.0/Data-Bool.html#v:False">False</a></code>.</li><li><strong>v</strong>: Left singular vectors. If <code>full_matrices</code> is <code><ahref="../base-4.8.2.0/Data-Bool.html#v:False">False</a></code> then shape is
`[..., N, N]`. If <code>full_matrices</code> is <code><ahref="../base-4.8.2.0/Data-Bool.html#v:True">True</a></code> then shape is `[..., N, P]`.
Undefined if <code>compute_uv</code> is false.</li></ul></td></tr></table></div><divclass="doc"><p>Computes the singular value decompositions of one or more matrices.</p><p>Computes the SVD of each inner matrix in <code>input</code> such that
# s is a tensor of singular values for each matrix.
# u is the tensor containing of left singular vectors for each matrix.
# v is the tensor containing of right singular vectors for each matrix.
s, u, v = svd(a)
s, _, _ = svd(a, compute_uv=False)
```</p></div></div><divclass="top"><pclass="src"><aname="v:matrixSolveLs"class="def">matrixSolveLs</a><ahref="src/TensorFlow-GenOps-Core.html#matrixSolveLs"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>matrix</strong>: Shape is `[..., M, N]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>rhs</strong>: Shape is `[..., M, K]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a></td><tdclass="doc"><p><strong>l2_regularizer</strong>: Scalar tensor.</p><p><code>compatibility(numpy)
Equivalent to np.linalg.lstsq
</code>end_compatibility</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Shape is `[..., N, K]`.</p></td></tr></table></div><divclass="doc"><p>Solves one or more linear least-squares problems.</p><p><code>matrix</code> is a tensor of shape `[..., M, N]` whose inner-most 2 dimensions
form matrices of size `[M, N]`. Rhs is a tensor of shape `[..., M, K]`.
The output is a tensor shape `[..., N, K]` where each output matrix solves
each of the equations matrix[..., :, :] * output[..., :, :] = rhs[..., :, :]
in the least squares sense.</p><p>matrix and right-hand sides in the batch:</p><p><code>matrix</code>=\(A in Re^{m times n}\),
<code>rhs</code>=\(B in Re^{m times k}\),
<code>output</code>=\(X in Re^{n times k}\),
<code>l2_regularizer</code>=\(lambda\).</p><p>If <code>fast</code> is <code><ahref="../base-4.8.2.0/Data-Bool.html#v:True">True</a></code>, then the solution is computed by solving the normal
equations using Cholesky decomposition. Specifically, if \(m ge n\) then
\(X = (A^T A + lambda I)^{-1} A^T B\), which solves the least-squares
problem \(X = mathrm{argmin}_{Z in Re^{n times k}} ||A Z - B||_F^2 +
lambda ||Z||_F^2\). If \(m lt n\) then <code>output</code> is computed as
\(X = A^T (A A^T + lambda I)^{-1} B\), which (for \(lambda = 0\)) is the
minimum-norm solution to the under-determined linear system, i.e.
\(X = mathrm{argmin}_{Z in Re^{n times k}} ||Z||_F^2 \), subject to
\(A Z = B\). Notice that the fast path is only numerically stable when
\(A\) is numerically full rank and has a condition number
\(mathrm{cond}(A) lt frac{1}{sqrt{epsilon_{mach}}}\) or\(lambda\) is
sufficiently large.</p><p>If <code>fast</code> is <code><ahref="../base-4.8.2.0/Data-Bool.html#v:False">False</a></code> an algorithm based on the numerically robust complete
orthogonal decomposition is used. This computes the minimum-norm
least-squares solution, even when \(A\) is rank deficient. This path is
typically 6-7 times slower than the fast path. If <code>fast</code> is <code><ahref="../base-4.8.2.0/Data-Bool.html#v:False">False</a></code> then
<code>l2_regularizer</code> is ignored.</p></div></div><divclass="top"><pclass="src"><aname="v:pack"class="def">pack</a><ahref="src/TensorFlow-GenOps-Core.html#pack"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t]</td><tdclass="doc"><p><strong>values</strong>: Must be of same shape and type.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The packed tensor.</p></td></tr></table></div><divclass="doc"><p>Packs a list of <code>N</code> rank-<code>R</code> tensors into one rank-`(R+1)` tensor.</p><p>Packs the <code>N</code> tensors in <code>values</code> into a tensor with rank one higher than each
```</p><p>This is the opposite of <code><ahref="TensorFlow-GenOps-Core.html#v:unpack">unpack</a></code>.</p></div></div><divclass="top"><pclass="src"><aname="v:barrierClose"class="def">barrierClose</a><ahref="src/TensorFlow-GenOps-Core.html#barrierClose"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a barrier.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Closes the given barrier.</p><p>This operation signals that no more new elements will be inserted in the
given barrier. Subsequent InsertMany that try to introduce a new key will fail.
Subsequent InsertMany operations that just add missing components to already
existing elements will continue to succeed. Subsequent TakeMany operations will
continue to succeed if sufficient completed elements remain in the barrier.
Subsequent TakeMany operations that would block will fail immediately.</p></div></div><divclass="top"><pclass="src"><aname="v:selfAdjointEigV2"class="def">selfAdjointEigV2</a><ahref="src/TensorFlow-GenOps-Core.html#selfAdjointEigV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> input of shape `[N, N]`.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p>(<strong>e</strong>, <strong>v</strong>)</p><ul><li><strong>e</strong>: Eigenvalues. Shape is `[N]`.</li><li><strong>v</strong>: Eigenvectors. Shape is `[N, N]`.</li></ul></td></tr></table></div><divclass="doc"><p>Computes the eigen decomposition of one or more square self-adjoint matrices.</p><p>Computes the eigenvalues and (optionally) eigenvectors of each inner matrix in
<code>input</code> such that `input[..., :, :] = v[..., :, :] * diag(e[..., :])`.</p><p>```prettyprint
# a is a tensor.
# e is a tensor of eigenvalues.
# v is a tensor of eigenvectors.
e, v = self_adjoint_eig(a)
e = self_adjoint_eig(a, compute_v=False)
```</p></div></div><divclass="top"><pclass="src"><aname="v:scatterSub"class="def">scatterSub</a><ahref="src/TensorFlow-GenOps-Core.html#scatterSub"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ref</strong>: Should be from a <code>Variable</code> node.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>indices</strong>: A tensor of indices into the first dimension of <code>ref</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>updates</strong>: A tensor of updated values to subtract from <code>ref</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output_ref</strong>: = Same as <code>ref</code>. Returned as a convenience for operations that want
to use the updated values after the update is done.</p></td></tr></table></div><divclass="doc"><p>Subtracts sparse updates to a variable reference.</p><p># Scalar indices
ref[indices, ...] -= updates[...]</p><p># Vector indices (for each i)
ref[indices[i], ...] -= updates[i, ...]</p><p># High rank indices (for each i, ..., j)
ref[indices[i, ..., j], ...] -= updates[i, ..., j, ...]</p><p>This operation outputs <code>ref</code> after the update is done.
This makes it easier to chain operations that need to use the reset value.</p><p>Duplicate entries are handled correctly: if multiple <code>indices</code> reference
the same location, their (negated) contributions add.</p><p>Requires `updates.shape = indices.shape + ref.shape[1:]`.</p><p><ahref="div">style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;"</a>
<ahref="/div">/div</a></p></div></div><divclass="top"><pclass="src"><aname="v:selfAdjointEig"class="def">selfAdjointEig</a><ahref="src/TensorFlow-GenOps-Core.html#selfAdjointEig"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Shape is `[..., M, M]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Shape is `[..., M+1, M]`.</p></td></tr></table></div><divclass="doc"><p>Computes the Eigen Decomposition of a batch of square self-adjoint matrices.</p><p>The input is a tensor of shape `[..., M, M]` whose inner-most 2 dimensions
form square matrices, with the same constraints as the single matrix
SelfAdjointEig.</p><p>The result is a [..., M+1, M] matrix with [..., 0,:] containing the
eigenvalues, and subsequent [...,1:, :] containing the eigenvectors.</p></div></div><divclass="top"><pclass="src"><aname="v:stopGradient"class="def">stopGradient</a><ahref="src/TensorFlow-GenOps-Core.html#stopGradient"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Stops gradient computation.</p><p>When executed in a graph, this op outputs its input tensor as-is.</p><p>When building ops to compute gradients, this op prevents the contribution of
its inputs to be taken into account. Normally, the gradient generator adds ops
to a graph to compute the derivatives of a specified <code>loss</code> by recursively
finding out inputs that contributed to its computation. If you insert this op
in the graph it inputs are masked from the gradient generator. They are not
taken into account for computing gradients.</p><p>This is useful any time you want to compute a value with TensorFlow but need
to pretend that the value was a constant. Some examples include:</p><ul><li>The *EM* algorithm where the *M-step* should not involve backpropagation
through the output of the *E-step*.</li><li>Contrastive divergence training of Boltzmann machines where, when
differentiating the energy function, the training must not backpropagate
through the graph that generated the samples from the model.</li><li>Adversarial training, where no backprop should happen through the adversarial
of the input Tensor to reduce across. For vectors, use dimension = 0.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Returns the index with the largest value across dimensions of a tensor.</p></div></div><divclass="top"><pclass="src"><aname="v:choleskyGrad"class="def">choleskyGrad</a><ahref="src/TensorFlow-GenOps-Core.html#choleskyGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>l</strong>: Output of batch Cholesky algorithm l = cholesky(A). Shape is `[..., M, M]`.
Algorithm depends only on lower triangular part of the innermost matrices of
this tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>grad</strong>: df/dl where f is some scalar function. Shape is `[..., M, M]`.
Algorithm depends only on lower triangular part of the innermost matrices of
this tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Symmetrized version of df/dA . Shape is `[..., M, M]`</p></td></tr></table></div><divclass="doc"><p>Computes the reverse mode backpropagated gradient of the Cholesky algorithm.</p><p>For an explanation see "Differentiation of the Cholesky algorithm" by
Iain Murray <ahref="http://arxiv.org/abs/1602.07527">http://arxiv.org/abs/1602.07527</a>.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseReshape"class="def">sparseReshape</a><ahref="src/TensorFlow-GenOps-Core.html#sparseReshape"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>input_indices</strong>: 2-D. `N x R_in` matrix with the indices of non-empty values in a
SparseTensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>input_shape</strong>: 1-D. <code>R_in</code> vector with the input SparseTensor's dense shape.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>new_shape</strong>: 1-D. <code>R_out</code> vector with the requested new dense shape.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p>(<strong>output_indices</strong>, <strong>output_shape</strong>)</p><ul><li><strong>output_indices</strong>: 2-D. `N x R_out` matrix with the updated indices of non-empty
values in the output SparseTensor.</li><li><strong>output_shape</strong>: 1-D. <code>R_out</code> vector with the full dense shape of the output
SparseTensor. This is the same as <code>new_shape</code> but with any -1 dimensions
filled in.</li></ul></td></tr></table></div><divclass="doc"><p>Reshapes a SparseTensor to represent values in a new dense shape.</p><p>This operation has the same semantics as reshape on the represented dense
tensor. The <code>input_indices</code> are recomputed based on the requested <code>new_shape</code>.</p><p>If one component of <code>new_shape</code> is the special value -1, the size of that
dimension is computed so that the total dense size remains constant. At
most one component of <code>new_shape</code> can be -1. The number of dense elements
implied by <code>new_shape</code> must be the same as the number of dense elements
originally implied by <code>input_shape</code>.</p><p>Reshaping does not affect the order of values in the SparseTensor.</p><p>If the input tensor has rank <code>R_in</code> and <code>N</code> non-empty values, and <code>new_shape</code>
has length <code>R_out</code>, then <code>input_indices</code> has shape `[N, R_in]`,
<code>input_shape</code> has length <code>R_in</code>, <code>output_indices</code> has shape `[N, R_out]`, and
<code>output_shape</code> has length <code>R_out</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseApplyAdadelta"class="def">sparseApplyAdadelta</a><ahref="src/TensorFlow-GenOps-Core.html#sparseApplyAdadelta"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>accum</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>accum_update</strong>: : Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>lr</strong>: Learning rate. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>rho</strong>: Decay factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 t</td><tdclass="doc"><p><strong>epsilon</strong>: Constant factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v7 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v8 tindices</td><tdclass="doc"><p><strong>indices</strong>: A vector of indices into the first dimension of var and accum.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>out</strong>: Same as "var".</p></td></tr></table></div><divclass="doc"><p>var: Should be from a Variable().</p></div></div><divclass="top"><pclass="src"><aname="v:dilation2DBackpropFilter"class="def">dilation2DBackpropFilter</a><ahref="src/TensorFlow-GenOps-Core.html#dilation2DBackpropFilter"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFl
components not set) in the barrier.</p></td></tr></table></div><divclass="doc"><p>Computes the number of incomplete elements in the given barrier.</p></div></div><divclass="top"><pclass="src"><aname="v:fakeQuantWithMinMaxVars"class="def">fakeQuantWithMinMaxVars</a><ahref="src/TensorFlow-GenOps-Core.html#fakeQuantWithMinMaxVars"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>inputs</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>min</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>max</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>outputs</strong></p></td></tr></table></div><divclass="doc"><p>Fake-quantize the <code>inputs</code> tensor of type float and shape `[b, h, w, d]` via</p><p>global float scalars <code><ahref="../base-4.8.2.0/Data-Ord.html#v:min">min</a></code> and <code><ahref="../base-4.8.2.0/Data-Ord.html#v:max">max</a></code> to <code>outputs</code> tensor of same shape as
<code>inputs</code>.</p><dl><dt>min; max</dt><dd>is the clamping range for the <code>inputs</code> data. Op divides this range
into 255 steps (total of 256 values), then replaces each <code>inputs</code> value with the
closest of the quantized step values.</dd></dl><p>This operation has a gradient and thus allows for training <code><ahref="../base-4.8.2.0/Data-Ord.html#v:min">min</a></code> and <code><ahref="../base-4.8.2.0/Data-Ord.html#v:max">max</a></code> values.</p></div></div><divclass="top"><pclass="src"><aname="v:readVariableOp"class="def">readVariableOp</a><ahref="src/TensorFlow-GenOps-Core.html#readVariableOp"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ResourceHandle">ResourceHandle</a> dtype</td><tdclass="doc"><p><strong>resource</strong>: handle to the resource in which to store the variable.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype)</td><tdclass="doc"><p><strong>value</strong></p></td></tr></table></div><divclass="doc"><p>Reads the value of a variable.</p><p>The tensor returned by this operation is immutable.</p><p>The value returned by this operation is guaranteed to be influenced by all the
writes on which this operation depends directly or indirectly, and to not be
influenced by any of the writes which depend directly or indirectly on this
operation.</p></div></div><divclass="top"><pclass="src"><aname="v:fusedBatchNormGrad"class="def">fusedBatchNormGrad</a><ahref="src/TensorFlow-GenOps-Core.html#fusedBatchNormGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>y_backprop</strong>: A 4D Tensor for the gradient with respect to y.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>x</strong>: A 4D Tensor for input data.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>scale</strong>: A 1D Tensor for scaling factor, to scale the normalized x.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>reserve_space_1</strong>: A 1D Tensor for the computed batch mean, to be reused
in the gradient computation.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>reserve_space_2</strong>: A 1D Tensor for the computed batch variance (inverted variance
in the cuDNN case), to be used in the gradient computation.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p>(<strong>x_backprop</strong>, <strong>scale_backprop</strong>, <strong>offset_backprop</strong>, <strong>reserve_space_3</strong>, <strong>reserve_space_4</strong>)</p><ul><li><strong>x_backprop</strong>: A 4D Tensor for the gradient with respect to x.</li><li><strong>scale_backprop</strong>: A 1D Tensor for the gradient with respect to scale.</li><li><strong>offset_backprop</strong>: A 1D Tensor for the gradient with respect to offset.</li><li><strong>reserve_space_3</strong>: Unused placeholder to match the mean input in FusedBatchNorm.</li><li><strong>reserve_space_4</strong>: Unused placeholder to match the variance input
in FusedBatchNorm.</li></ul></td></tr></table></div><divclass="doc"><p>Gradient for batch normalization.</p><p>Note that the size of 4D Tensors are defined by either <ahref="NHWC.html">NHWC</a> or <ahref="NCHW.html">NCHW</a>.
The size of 1D Tensors matches the dimension C of the 4D Tensors.</p></div></div><divclass="top"><pclass="src"><aname="v:paddingFIFOQueue"class="def">paddingFIFOQueue</a><ahref="src/TensorFlow-GenOps-Core.html#paddingFIFOQueue"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>)</td><tdclass="doc"><p><strong>handle</strong>: The handle to the queue.</p></td></tr></table></div><divclass="doc"><p>A queue that produces elements in first-in first-out order.</p><p>Variable-size shapes are allowed by setting the corresponding shape dimensions
to 0 in the shape attr. In this case DequeueMany will pad up to the maximum
size of any given element in the minibatch. See below for details.</p></div></div><divclass="top"><pclass="src"><aname="v:matrixInverse"class="def">matrixInverse</a><ahref="src/TensorFlow-GenOps-Core.html#matrixInverse"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Shape is `[..., M, M]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Shape is `[..., M, M]`.</p><p><code>compatibility(numpy)
Equivalent to np.linalg.inv
</code>end_compatibility</p></td></tr></table></div><divclass="doc"><p>Computes the inverse of one or more square invertible matrices or their</p><p>adjoints (conjugate transposes).</p><p>The input is a tensor of shape `[..., M, M]` whose inner-most 2 dimensions
form square matrices. The output is a tensor of the same shape as the input
containing the inverse for all input submatrices `[..., :, :]`.</p><p>The op uses LU decomposition with partial pivoting to compute the inverses.</p><p>If a matrix is not invertible there is no guarantee what the op does. It
may detect the condition and raise an exception or it may simply return a
garbage result.</p></div></div><divclass="top"><pclass="src"><aname="v:audioSummaryV2"class="def">audioSummaryV2</a><ahref="src/TensorFlow-GenOps-Core.html#audioSummaryV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>tag</strong>: Scalar. Used to build the <code>tag</code> attribute of the summary values.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>tensor</strong>: 2-D of shape `[batch_size, frames]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>sample_rate</strong>: The sample rate of the signal in hertz.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>summary</strong>: Scalar. Serialized <code>Summary</code> protocol buffer.</p></td></tr></table></div><divclass="doc"><p>Outputs a <code>Summary</code> protocol buffer with audio.</p><p>The summary has up to <code>max_outputs</code> summary values containing audio. The
audio is built from <code>tensor</code> which must be 3-D with shape `[batch_size,
frames, channels]` or 2-D with shape `[batch_size, frames]`. The values are
assumed to be in the range of `[-1.0, 1.0]` with a sample rate of <code>sample_rate</code>.</p><p>The <code>tag</code> argument is a scalar <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> of type <code>string</code>. It is used to
build the <code>tag</code> of the summary values:</p><ul><li>If <code>max_outputs</code> is 1, the summary value tag is '*tag*/audio'.</li><li>If <code>max_outputs</code> is greater than 1, the summary value tags are
generated sequentially as '*tag*/audio/0', '*tag*/audio/1', etc.</li></ul></div></div><divclass="top"><pclass="src"><aname="v:matrixDeterminant"class="def">matrixDeterminant</a><ahref="src/TensorFlow-GenOps-Core.html#matrixDeterminant"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Shape is `[..., M, M]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Shape is `[...]`.</p></td></tr></table></div><divclass="doc"><p>Computes the determinant of one ore more square matrices.</p><p>The input is a tensor of shape `[..., M, M]` whose inner-most 2 dimensions
form square matrices. The output is a tensor containing the determinants
for all input submatrices `[..., :, :]`.</p></div></div><divclass="top"><pclass="src"><aname="v:writeFile"class="def">writeFile</a><ahref="src/TensorFlow-GenOps-Core.html#writeFile"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>filename</strong>: scalar. The name of the file to which we write the contents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>contents</strong>: scalar. The content to be written to the output file.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Writes contents to the file at input filename. Creates file if not existing.</p></div></div><divclass="top"><pclass="src"><aname="v:quantizedConcat"class="def">quantizedConcat</a><ahref="src/TensorFlow-GenOps-Core.html#quantizedConcat"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>concat_dim</strong>: 0-D. The dimension along which to concatenate. Must be in the
range [0, rank(values)).</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t]</td><tdclass="doc"><p><strong>values</strong>: The <code>N</code> Tensors to concatenate. Their ranks and types must match,
and their sizes must match in all dimensions except <code>concat_dim</code>.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]</td><tdclass="doc"><p><strong>input_mins</strong>: The minimum scalar values for each of the input tensors.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]</td><tdclass="doc"><p><strong>input_maxes</strong>: The maximum scalar values for each of the input tensors.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>output</strong>, <strong>output_min</strong>, <strong>output_max</strong>)</p><ul><li><strong>output</strong>: A <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> with the concatenation of values stacked along the
<code>concat_dim</code> dimension. This tensor's shape matches that of <code>values</code> except
in <code>concat_dim</code> where it has the sum of the sizes.</li><li><strong>output_min</strong>: The float value that the minimum quantized output value represents.</li><li><strong>output_max</strong>: The float value that the maximum quantized output value represents.</li></ul></td></tr></table></div><divclass="doc"><p>Concatenates quantized tensors along one dimension.</p></div></div><divclass="top"><pclass="src"><aname="v:varHandleOp"class="def">varHandleOp</a><ahref="src/TensorFlow-GenOps-Core.html#varHandleOp"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:Shape">Shape</a></td><tdclass="doc"><p><strong>shape</strong>: The (possibly partially specified) shape of this variable.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ResourceHandle">ResourceHandle</a> dtype)</td><tdclass="doc"><p><strong>resource</strong></p></td></tr></table></div><divclass="doc"><p>Creates a handle to a Variable resource.</p></div></div><divclass="top"><pclass="src"><aname="v:stridedSliceAssign"class="def">stridedSliceAssign</a><ahref="src/TensorFlow-GenOps-Core.html#stridedSliceAssign"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> index, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` index)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ref</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 index</td><tdclass="doc"><p><strong>begin</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 index</td><tdclass="doc"><p><strong>end</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 index</td><tdclass="doc"><p><strong>strides</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>value</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output_ref</strong></p></td></tr></table></div><divclass="doc"><p>Assign <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code> to the sliced l-value reference of <code>ref</code>.</p><p>The values of <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code> are assigned to the positions in the variable
<code>ref</code> that are selected by the slice parameters. The slice parameters
`begin, <code>end</code>, <code>strides</code>, etc. work exactly as in <code>StridedSlice</code>.</p><p>NOTE this op currently does not support broadcasting and so <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>'s
shape must be exactly the shape produced by the slice of <code>ref</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:varIsInitializedOp"class="def">varIsInitializedOp</a><ahref="src/TensorFlow-GenOps-Core.html#varIsInitializedOp"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ResourceHandle">ResourceHandle</a> dtype</td><tdclass="doc"><p><strong>resource</strong>: the input resource handle.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a>)</td><tdclass="doc"><p><strong>is_initialized</strong>: a scalar boolean which is true if the variable has been
initialized.</p></td></tr></table></div><divclass="doc"><p>Checks whether a resource handle-based variable has been initialized.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseApplyRMSProp"class="def">sparseApplyRMSProp</a><ahref="src/TensorFlow-GenOps-Core.html#sparseApplyRMSProp"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ms</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>mom</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>lr</strong>: Scaling factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>rho</strong>: Decay rate. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 t</td><tdclass="doc"><p><strong>momentum</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v7 t</td><tdclass="doc"><p><strong>epsilon</strong>: Ridge term. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v8 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v9 tindices</td><tdclass="doc"><p><strong>indices</strong>: A vector of indices into the first dimension of var, ms and mom.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>out</strong>: Same as "var".</p></td></tr></table></div><divclass="doc"><p>Update '*var' according to the RMSProp algorithm.</p><p>Note that in dense implementation of this algorithm, ms and mom will
update even if the grad is zero, but in this sparse implementation, ms
and mom will not update in iterations during which the grad is zero.</p><p>mean_square = decay * mean_square + (1-decay) * gradient ** 2
Delta = learning_rate * gradient / sqrt(mean_square + epsilon)</p><p>ms <- rho * ms_{t-1} + (1-rho) * grad * grad
mom <- momentum * mom_{t-1} + lr * grad / sqrt(ms + epsilon)
var <- var - mom</p></div></div><divclass="top"><pclass="src"><aname="v:batchCholesky"class="def">batchCholesky</a><ahref="src/TensorFlow-GenOps-Core.html#batchCholesky"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:tensorArrayGather"class="def">tensorArrayGather</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArrayGather"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>indices</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype)</td><tdclass="doc"><p><strong>value</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:readerRestoreState"class="def">readerRestoreState</a><ahref="src/TensorFlow-GenOps-Core.html#readerRestoreState"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>reader_handle</strong>: Handle to a Reader.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>state</strong>: Result of a ReaderSerializeState of a Reader with type
matching reader_handle.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Restore a reader to a previously saved state.</p><p>Not all Readers support being restored, so this can produce an
Unimplemented error.</p></div></div><divclass="top"><pclass="src"><aname="v:sqrtGrad"class="def">sqrtGrad</a><ahref="src/TensorFlow-GenOps-Core.html#sqrtGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Computes the gradient for the sqrt of <code>x</code> wrt its input.</p><p>Specifically, `grad = dy * 0.5 / y`, where `y = sqrt(x)`, and <code>dy</code>
is the corresponding input gradient.</p></div></div><divclass="top"><pclass="src"><aname="v:split"class="def">split</a><ahref="src/TensorFlow-GenOps-Core.html#split"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_split</strong>: The number of ways to split. Must evenly divide
`value.shape[split_dim]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>split_dim</strong>: 0-D. The dimension along which to split. Must be in the range
`[0, rank(value))`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>value</strong>: The tensor to split.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t]</td><tdclass="doc"><p><strong>output</strong>: They are identically shaped tensors, whose shape matches that of <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>
except along <code>split_dim</code>, where their sizes are
`values.shape[split_dim] / num_split`.</p></td></tr></table></div><divclass="doc"><p>Splits a tensor into <code>num_split</code> tensors along one dimension.</p></div></div><divclass="top"><pclass="src"><aname="v:textLineReader"class="def">textLineReader</a><ahref="src/TensorFlow-GenOps-Core.html#textLineReader"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>)</td><tdclass="doc"><p><strong>reader_handle</strong>: The handle to reference the Reader.</p></td></tr></table></div><divclass="doc"><p>A Reader that outputs the lines of a file delimited by '\n'.</p></div></div><divclass="top"><pclass="src"><aname="v:matrixBandPart"class="def">matrixBandPart</a><ahref="src/TensorFlow-GenOps-Core.html#matrixBandPart"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Rank <code>k</code> tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_lower</strong>: 0-D tensor. Number of subdiagonals to keep. If negative, keep entire
lower triangle.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_upper</strong>: 0-D tensor. Number of superdiagonals to keep. If negative, keep
entire upper triangle.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>band</strong>: Rank <code>k</code> tensor of the same shape as input. The extracted banded tensor.</p></td></tr></table></div><divclass="doc"><p>Copy a tensor setting everything outside a central band in each innermost matrix</p><p>to zero.</p><p>The <code>band</code> part is computed as follows:
Assume <code>input</code> has <code>k</code> dimensions `[I, J, K, ..., M, N]`, then the output is a
tensor with the same shape where</p><p>`band[i, j, k, ..., m, n] = in_band(m, n) * input[i, j, k, ..., m, n]`.</p><p>The indicator function</p><p>`in_band(m, n) = (num_lower < 0 || (m-n) <= num_lower)) &&
```</p></div></div><divclass="top"><pclass="src"><aname="v:queueClose"class="def">queueClose</a><ahref="src/TensorFlow-GenOps-Core.html#queueClose"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a queue.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Closes the given queue.</p><p>This operation signals that no more elements will be enqueued in the
given queue. Subsequent Enqueue(Many) operations will fail.
Subsequent Dequeue(Many) operations will continue to succeed if
sufficient elements remain in the queue. Subsequent Dequeue(Many)
operations that would block will fail immediately.</p></div></div><divclass="top"><pclass="src"><aname="v:mergeV2Checkpoints"class="def">mergeV2Checkpoints</a><ahref="src/TensorFlow-GenOps-Core.html#mergeV2Checkpoints"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>checkpoint_prefixes</strong>: prefixes of V2 checkpoints to merge.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>destination_prefix</strong>: scalar. The desired final prefix. Allowed to be the same
as one of the checkpoint_prefixes.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>V2 format specific: merges the metadata files of sharded checkpoints. The</p><p>result is one logical checkpoint, with one physical metadata file and renamed
data files.</p><p>Intended for "grouping" multiple checkpoints in a sharded checkpoint setup.</p><p>If delete_old_dirs is true, attempts to delete recursively the dirname of each
path in the input checkpoint_prefixes. This is useful when those paths are non
user-facing temporary locations.</p></div></div><divclass="top"><pclass="src"><aname="v:barrierReadySize"class="def">barrierReadySize</a><ahref="src/TensorFlow-GenOps-Core.html#barrierReadySize"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a barrier.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>)</td><tdclass="doc"><p><strong>size</strong>: The number of complete elements (i.e. those with all of their value
components set) in the barrier.</p></td></tr></table></div><divclass="doc"><p>Computes the number of complete elements in the given barrier.</p></div></div><divclass="top"><pclass="src"><aname="v:randomShuffleQueue"class="def">randomShuffleQueue</a><ahref="src/TensorFlow-GenOps-Core.html#randomShuffleQueue"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>)</td><tdclass="doc"><p><strong>handle</strong>: The handle to the queue.</p></td></tr></table></div><divclass="doc"><p>A queue that randomizes the order of elements.</p></div></div><divclass="top"><pclass="src"><aname="v:notEqual"class="def">notEqual</a><ahref="src/TensorFlow-GenOps-Core.html#notEqual"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a>, <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns the truth value of (x != y) element-wise.</p><ul><li>NOTE*: <code>NotEqual</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:nonMaxSuppression"class="def">nonMaxSuppression</a><ahref="src/TensorFlow-GenOps-Core.html#nonMaxSuppression"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>boxes</strong>: A 2-D float tensor of shape `[num_boxes, 4]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>scores</strong>: A 1-D float tensor of shape `[num_boxes]` representing a single
score corresponding to each box (each row of boxes).</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>max_output_size</strong>: A scalar integer tensor representing the maximum number of
boxes to be selected by non max suppression.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>selected_indices</strong>: A 1-D integer tensor of shape `[M]` representing the selected
indices from the boxes tensor, where `M <= max_output_size`.</p></td></tr></table></div><divclass="doc"><p>Greedily selects a subset of bounding boxes in descending order of score,</p><p>pruning away boxes that have high intersection-over-union (IOU) overlap
with previously selected boxes. Bounding boxes are supplied as
[y1, x1, y2, x2], where (y1, x1) and (y2, x2) are the coordinates of any
diagonal pair of box corners and the coordinates can be provided as normalized
(i.e., lying in the interval [0, 1]) or absolute. Note that this algorithm
is agnostic to where the origin is in the coordinate system. Note that this
algorithm is invariant to orthogonal transformations and translations
of the coordinate system; thus translating or reflections of the coordinate
system result in the same boxes being selected by the algorithm.</p><p>The output of this operation is a set of integers indexing into the input
collection of bounding boxes representing the selected boxes. The bounding
box coordinates corresponding to the selected indices can then be obtained
using the `tf.gather operation`. For example:</p><p>selected_indices = tf.image.non_max_suppression(
boxes, scores, max_output_size, iou_threshold)
selected_boxes = tf.gather(boxes, selected_indices)</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArrayWrite"class="def">tensorArrayWrite</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArrayWrite"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>index</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>value</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>flow_out</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:quantizeAndDequantize"class="def">quantizeAndDequantize</a><ahref="src/TensorFlow-GenOps-Core.html#quantizeAndDequantize"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Tensor to quantize and then dequantize.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Quantizes then dequantizes a tensor.</p><p>This op simulates the precision loss from the quantized forward pass by:
1. Quantizing the tensor to fixed point numbers, which should match the target
quantization method when it is used in inference.
2. Dequantizing it back to floating point numbers for the following ops, most
likely matmul.</p><p>There are different ways to quantize. This version does not use the full range
of the output type, choosing to elide the lowest possible value for symmetry
(e.g., output range is -127 to 127, not -128 to 127 for signed 8 bit
quantization), so that 0.0 maps to 0.</p><p>To perform this op, we first find the range of values in our tensor. The range
we use is always centered on 0, so we find m such that</p><ol><li>m = max(abs(input_min), abs(input_max)) if range_given is true,</li><li>m = max(max(abs(min_elem(input)), abs(max_elem(input))) otherwise.</li></ol><p>Our input tensor range is then [-m, m].</p><p>Next, we choose our fixed-point quantization buckets, [min_fixed, max_fixed].
If signed_input is true, this is</p><dl><dt>min_fixed, max_fixed </dt><dd>=</dd><dt>-(1 << (num_bits - 1) - 1), (1 << (num_bits - 1)) - 1</dt><dd>.</dd></dl><p>Otherwise, if signed_input is false, the fixed-point range is</p><dl><dt>min_fixed, max_fixed</dt><dd>= [0, (1 << num_bits) - 1].</dd></dl><p>From this we compute our scaling factor, s:</p><p>s = (max_fixed - min_fixed) / (2 * m).</p><p>Now we can quantize and dequantize the elements of our tensor. An element e
is transformed into e':</p><p>e' = (e * s).round_to_nearest() / s.</p><p>Note that we have a different number of buckets in the signed vs. unsigned
cases. For example, if num_bits == 8, we get 254 buckets in the signed case
vs. 255 in the unsigned case.</p><p>For example, suppose num_bits = 8 and m = 1. Then</p><dl><dt>min_fixed, max_fixed</dt><dd>= [-127, 127], and
s = (127 + 127) / 2 = 127.</dd></dl><p>Given the vector {-1, -0.5, 0, 0.3}, this is quantized to
{-127, -63, 0, 38}, and dequantized to {-1, -63.0<em>127, 0, 38.0</em>127}.</p></div></div><divclass="top"><pclass="src"><aname="v:readerRead"class="def">readerRead</a><ahref="src/TensorFlow-GenOps-Core.html#readerRead"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>reader_handle</strong>: Handle to a Reader.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>queue_handle</strong>: Handle to a Queue, with string work items.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>)</td><tdclass="doc"><p>(<strong>key</strong>, <strong>value</strong>)</p><ul><li><strong>key</strong>: A scalar.</li><li><strong>value</strong>: A scalar.</li></ul></td></tr></table></div><divclass="doc"><p>Returns the next record (key, value pair) produced by a Reader.</p><p>Will dequeue from the input queue if necessary (e.g. when the
Reader needs to start reading from a new file since it has finished
with the previous file).</p></div></div><divclass="top"><pclass="src"><aname="v:matrixTriangularSolve"class="def">matrixTriangularSolve</a><ahref="src/TensorFlow-GenOps-Core.html#matrixTriangularSolve"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>matrix</strong>: Shape is `[..., M, M]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>rhs</strong>: Shape is `[..., M, K]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Shape is `[..., M, K]`.</p></td></tr></table></div><divclass="doc"><p>Solves systems of linear equations with upper or lower triangular matrices by</p><p>backsubstitution.</p><p><code>matrix</code> is a tensor of shape `[..., M, M]` whose inner-most 2 dimensions form
square matrices. If <code>lower</code> is <code><ahref="../base-4.8.2.0/Data-Bool.html#v:True">True</a></code> then the strictly upper triangular part
of each inner-most matrix is assumed to be zero and not accessed.
If <code>lower</code> is False then the strictly lower triangular part of each inner-most
matrix is assumed to be zero and not accessed.
<code>rhs</code> is a tensor of shape `[..., M, K]`.</p><p>The output is a tensor of shape `[..., M, K]`. If <code>adjoint</code> is
<code><ahref="../base-4.8.2.0/Data-Bool.html#v:True">True</a></code> then the innermost matrices in output` satisfy matrix equations
If <code>adjoint</code> is <code><ahref="../base-4.8.2.0/Data-Bool.html#v:False">False</a></code> then the strictly then the innermost matrices in
<code>output</code> satisfy matrix equations
`adjoint(matrix[..., i, k]) * output[..., k, j] = rhs[..., i, j]`.</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArraySplitV2"class="def">tensorArraySplitV2</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArraySplitV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a TensorArray.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>value</strong>: The concatenated tensor to write to the TensorArray.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>lengths</strong>: The vector of lengths, how to split the rows of value into the
TensorArray.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong>: A float scalar that enforces proper chaining of operations.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_out</strong>: A float scalar that enforces proper chaining of operations.</p></td></tr></table></div><divclass="doc"><p>Split the data from the input value into TensorArray elements.</p><p>Assuming that <code>lengths</code> takes on values</p><p>```(n0, n1, ..., n(T-1))```</p><p>and that <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code> has shape</p><p>```(n0 + n1 + ... + n(T-1) x d0 x d1 x ...)```,</p><p>this splits values into a TensorArray with T tensors.</p><p>TensorArray index t will be the subtensor of values with starting position</p><p>```(n0 + n1 + ... + n(t-1), 0, 0, ...)```</p><p>and having size</p><p>```nt x d0 x d1 x ...```</p></div></div><divclass="top"><pclass="src"><aname="v:restore"class="def">restore</a><ahref="src/TensorFlow-GenOps-Core.html#restore"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dt</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>file_pattern</strong>: Must have a single element. The pattern of the files from
which we read the tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>tensor_name</strong>: Must have a single element. The name of the tensor to be
restored.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dt</td><tdclass="doc"><p><strong>tensor</strong>: The restored tensor.</p></td></tr></table></div><divclass="doc"><p>Restores a tensor from checkpoint files.</p><p>Reads a tensor stored in one or several files. If there are several files (for
instance because a tensor was saved as slices), <code>file_pattern</code> may contain
wildcard symbols (<code><ahref="../base-4.8.2.0/Prelude.html#v:-42-">*</a></code> and <code>?</code>) in the filename portion only, not in the
directory portion.</p><p>If a <code>file_pattern</code> matches several files, <code>preferred_shard</code> can be used to hint
in which file the requested tensor is likely to be found. This op will first
open the file at index <code>preferred_shard</code> in the list of matching files and try
to restore tensors from that file. Only if some tensors or tensor slices are
not found in that first file, then the Op opens all the files. Setting
<code>preferred_shard</code> to match the value passed as the <code>shard</code> input
of a matching <code>Save</code> Op may speed up Restore. This attribute only affects
performance, not correctness. The default value -1 means files are processed in
order.</p><p>See also <code>RestoreSlice</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:quantizedReluX"class="def">quantizedReluX</a><ahref="src/TensorFlow-GenOps-Core.html#quantizedReluX"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tinput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` tinput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` out_type)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 tinput</td><tdclass="doc"><p><strong>features</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>max_value</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>min_features</strong>: The float value that the lowest quantized value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>max_features</strong>: The float value that the highest quantized value represents.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>activations</strong>, <strong>min_activations</strong>, <strong>max_activations</strong>)</p><ul><li><strong>activations</strong>: Has the same output shape as "features".</li><li><strong>min_activations</strong>: The float value that the lowest quantized value represents.</li><li><strong>max_activations</strong>: The float value that the highest quantized value represents.</li></ul></td></tr></table></div><divclass="doc"><p>Computes Quantized Rectified Linear X: `min(max(features, 0), max_value)`</p></div></div><divclass="top"><pclass="src"><aname="v:accumulatorTakeGradient"class="def">accumulatorTakeGradient</a><ahref="src/TensorFlow-GenOps-Core.html#accumulatorTakeGradient"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a
The op blocks until sufficient gradients have been accumulated.
If the accumulator has already aggregated more than num_required gradients, it
returns the average of the accumulated gradients.
Also automatically increments the recorded global_step in the accumulator by 1,
and resets the aggregate to 0.</p></div></div><divclass="top"><pclass="src"><aname="v:floorMod"class="def">floorMod</a><ahref="src/TensorFlow-GenOps-Core.html#floorMod"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns element-wise remainder of division. When `x < 0` xor `y < 0` is</p><p>true, this follows Python semantics in that the result here is consistent
with a flooring divide. E.g. `floor(x / y) * y + mod(x, y) = x`.</p><ul><li>NOTE*: <code>FloorMod</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:matchingFiles"class="def">matchingFiles</a><ahref="src/TensorFlow-GenOps-Core.html#matchingFiles"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>pattern</strong>: A (scalar) shell wildcard pattern.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>filenames</strong>: A vector of matching filenames.</p></td></tr></table></div><divclass="doc"><p>Returns the set of files matching a pattern.</p><p>Note that this routine only supports wildcard characters in the
basename portion of the pattern, not in the directory portion.</p></div></div><divclass="top"><pclass="src"><aname="v:maxPool"class="def">maxPool</a><ahref="src/TensorFlow-GenOps-Core.html#maxPool"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: 4-D input to pool over.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The max pooled output tensor.</p></td></tr></table></div><divclass="doc"><p>Performs max pooling on the input.</p></div></div><divclass="top"><pclass="src"><aname="v:computeAccidentalHits"class="def">computeAccidentalHits</a><ahref="src/TensorFlow-GenOps-Core.html#computeAccidentalHits"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_true</strong>: Number of true labels per context.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>true_classes</strong>: The true_classes output of UnpackSparseLabels.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sampled_candidates</strong>: The sampled_candidates output of CandidateSampler.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>indices</strong>, <strong>ids</strong>, <strong>weights</strong>)</p><ul><li><strong>indices</strong>: A vector of indices corresponding to rows of true_candidates.</li><li><strong>ids</strong>: A vector of IDs of positions in sampled_candidates that match a true_label
for the row with the corresponding index in indices.</li><li><strong>weights</strong>: A vector of the same length as indices and ids, in which each element
is -FLOAT_MAX.</li></ul></td></tr></table></div><divclass="doc"><p>Computes the ids of the positions in sampled_candidates that match true_labels.</p><p>When doing log-odds NCE, the result of this op should be passed through a
SparseToDense op, then added to the logits of the sampled candidates. This has
the effect of <code>removing</code> the sampled labels that match the true labels by
making the classifier sure that they are sampled labels.</p></div></div><divclass="top"><pclass="src"><aname="v:deserializeManySparse"class="def">deserializeManySparse</a><ahref="src/TensorFlow-GenOps-Core.html#deserializeManySparse"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>serialized_sparse</strong>: 2-D, The <code>N</code> serialized <code>SparseTensor</code> objects.
Must have 3 columns.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p>(<strong>sparse_indices</strong>, <strong>sparse_values</strong>, <strong>sparse_shape</strong>)</p><ul><li><strong>sparse_indices</strong></li><li><strong>sparse_values</strong></li><li><strong>sparse_shape</strong></li></ul></td></tr></table></div><divclass="doc"><p>Deserialize and concatenate <code>SparseTensors</code> from a serialized minibatch.</p><p>The input <code>serialized_sparse</code> must be a string matrix of shape `[N x 3]` where
<code>N</code> is the minibatch size and the rows correspond to packed outputs of
<code>SerializeSparse</code>. The ranks of the original <code>SparseTensor</code> objects
must all match. When the final <code>SparseTensor</code> is created, it has rank one
higher than the ranks of the incoming <code>SparseTensor</code> objects
(they have been concatenated along a new row dimension).</p><p>The output <code>SparseTensor</code> object's shape values for all dimensions but the
first are the max across the input <code>SparseTensor</code> objects' shape values
for the corresponding dimensions. Its first shape value is <code>N</code>, the minibatch
size.</p><p>The input <code>SparseTensor</code> objects' indices are assumed ordered in
standard lexicographic order. If this is not the case, after this
step run <code>SparseReorder</code> to restore index ordering.</p><p>For example, if the serialized input is a `[2 x 3]` matrix representing two
original <code>SparseTensor</code> objects:</p><p>index = [ 0]
[10]
[20]
values = [1, 2, 3]
shape = [50]</p><p>and</p><p>index = [ 2]
[10]
values = [4, 5]
shape = [30]</p><p>then the final deserialized <code>SparseTensor</code> will be:</p><p>index = [0 0]
Both <code>image_height</code> and <code>image_width</code> need to be positive.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>boxes</strong>: A 2-D tensor of shape `[num_boxes, 4]`. The <code>i</code>-th row of the tensor
specifies the coordinates of a box in the `box_ind[i]` image and is specified
in normalized coordinates `[y1, x1, y2, x2]`. A normalized coordinate value of
<code>y</code> is mapped to the image coordinate at `y * (image_height - 1)`, so as the
`[0, 1]` interval of normalized image height is mapped to
`[0, image_height - 1] in image height coordinates. We do allow y1 > y2, in
which case the sampled crop is an up-down flipped version of the original
image. The width dimension is treated similarly. Normalized coordinates
outside the `[0, 1]` range are allowed, in which case we use
<code>extrapolation_value</code> to extrapolate the input image values.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>box_ind</strong>: A 1-D tensor of shape `[num_boxes]` with int32 values in `[0, batch)`.
The value of `box_ind[i]` specifies the image that the <code>i</code>-th box refers to.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>crop_size</strong>: A 1-D tensor of 2 elements, `size = [crop_height, crop_width]`. All
cropped image patches are resized to this size. The aspect ratio of the image
content is not preserved. Both <code>crop_height</code> and <code>crop_width</code> need to be
positive.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>crops</strong>: A 4-D tensor of shape `[num_boxes, crop_height, crop_width, depth]`.</p></td></tr></table></div><divclass="doc"><p>Extracts crops from the input image tensor and bilinearly resizes them (possibly</p><p>with aspect ratio change) to a common output size specified by <code>crop_size</code>. This
is more general than the <code>crop_to_bounding_box</code> op which extracts a fixed size
slice from the input image and does not allow resizing or aspect ratio change.</p><p>Returns a tensor with <code>crops</code> from the input <code>image</code> at positions defined at the
bounding box locations in <code>boxes</code>. The cropped boxes are all resized (with
bilinear interpolation) to a fixed `size = [crop_height, crop_width]`. The
result is a 4-D tensor `[num_boxes, crop_height, crop_width, depth]`.</p></div></div><divclass="top"><pclass="src"><aname="v:scatterUpdate"class="def">scatterUpdate</a><ahref="src/TensorFlow-GenOps-Core.html#scatterUpdate"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ref</strong>: Should be from a <code>Variable</code> node.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>indices</strong>: A tensor of indices into the first dimension of <code>ref</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>updates</strong>: A tensor of updated values to store in <code>ref</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output_ref</strong>: = Same as <code>ref</code>. Returned as a convenience for operations that want
to use the updated values after the update is done.</p></td></tr></table></div><divclass="doc"><p>Applies sparse updates to a variable reference.</p><p>This operation computes</p><p># Scalar indices
ref[indices, ...] = updates[...]</p><p># Vector indices (for each i)
ref[indices[i], ...] = updates[i, ...]</p><p># High rank indices (for each i, ..., j)
ref[indices[i, ..., j], ...] = updates[i, ..., j, ...]</p><p>This operation outputs <code>ref</code> after the update is done.
This makes it easier to chain operations that need to use the reset value.</p><p>If values in <code>ref</code> is to be updated more than once, because there are
duplicate entires in <code>indices</code>, the order at which the updates happen
for each value is undefined.</p><p>Requires `updates.shape = indices.shape + ref.shape[1:]`.</p><p><ahref="div">style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;"</a>
<ahref="/div">/div</a></p></div></div><divclass="top"><pclass="src"><aname="v:randomGamma"class="def">randomGamma</a><ahref="src/TensorFlow-GenOps-Core.html#randomGamma"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> s, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` s, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 s</td><tdclass="doc"><p><strong>shape</strong>: 1-D integer tensor. Shape of independent samples to draw from each
distribution described by the shape parameters given in alpha.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>alpha</strong>: A tensor in which each scalar is a "shape" parameter describing the
associated gamma distribution.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p><strong>output</strong>: A tensor with shape `shape + shape(alpha)`. Each slice
`[:, ..., :, i0, i1, ...iN]` contains the samples drawn for
`alpha[i0, i1, ...iN]`. The dtype of the output matches the dtype of alpha.</p></td></tr></table></div><divclass="doc"><p>Outputs random values from the Gamma distribution(s) described by alpha.</p><p>This op uses the algorithm by Marsaglia et al. to acquire samples via
transformation-rejection from pairs of uniform and normal random variables.
<code>keep_dims</code> is true, the rank of the tensor is reduced by 1 for each entry in
<code>reduction_indices</code>. If <code>keep_dims</code> is true, the reduced dimensions are
retained with length 1.</p></div></div><divclass="top"><pclass="src"><aname="v:readerNumRecordsProduced"class="def">readerNumRecordsProduced</a><ahref="src/TensorFlow-GenOps-Core.html#readerNumRecordsProduced"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>reader_handle</strong>: Handle to a Reader.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p><strong>records_produced</strong></p></td></tr></table></div><divclass="doc"><p>Returns the number of records this Reader has produced.</p><p>This is the same as the number of ReaderRead executions that have
succeeded.</p></div></div><divclass="top"><pclass="src"><aname="v:stackPop"class="def">stackPop</a><ahref="src/TensorFlow-GenOps-Core.html#stackPop"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> elem_type</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a stack.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> elem_type)</td><tdclass="doc"><p><strong>elem</strong>: The tensor that is popped from the top of the stack.</p></td></tr></table></div><divclass="doc"><p>Pop the element at the top of the stack.</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArrayScatterV2"class="def">tensorArrayScatterV2</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArrayScatterV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a TensorArray.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>indices</strong>: The locations at which to write the tensor elements.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>value</strong>: The concatenated tensor to write to the TensorArray.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong>: A float scalar that enforces proper chaining of operations.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_out</strong>: A float scalar that enforces proper chaining of operations.</p></td></tr></table></div><divclass="doc"><p>Scatter the data from the input value into specific TensorArray elements.</p><p><code>indices</code> must be a vector, its length must match the first dim of <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>.</p></div></div><divclass="top"><pclass="src"><aname="v:rGBToHSV"class="def">rGBToHSV</a><ahref="src/TensorFlow-GenOps-Core.html#rGBToHSV"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>images</strong>: 1-D or higher rank. RGB data to convert. Last dimension m
value of the pixels. The output is only well defined if the value in <code>images</code>
are in `[0,1]`.</p><p>`output[..., 0]` contains hue, `output[..., 1]` contains saturation, and
`output[..., 2]` contains value. All HSV values are in `[0,1]`. A hue of 0
corresponds to pure red, hue 1<em>3 is pure green, and 2</em>3 is pure blue.</p></div></div><divclass="top"><pclass="src"><aname="v:serializeManySparse"class="def">serializeManySparse</a><ahref="src/TensorFlow-GenOps-Core.html#serializeManySparse"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sparse_indices</strong>: 2-D. The <code>indices</code> of the minibatch <code>SparseTensor</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>sparse_values</strong>: 1-D. The <code>values</code> of the minibatch <code>SparseTensor</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sparse_shape</strong>: 1-D. The <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> of the minibatch <code>SparseTensor</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>serialized_sparse</strong></p></td></tr></table></div><divclass="doc"><p>Serialize an <code>N</code>-minibatch <code>SparseTensor</code> into an `[N, 3]` string <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code>.</p><p>The <code>SparseTensor</code> must have rank <code>R</code> greater than 1, and the first dimension
is treated as the minibatch dimension. Elements of the <code>SparseTensor</code>
must be sorted in increasing order of this first dimension. The serialized
<code>SparseTensor</code> objects going into each row of <code>serialized_sparse</code> will have
rank `R-1`.</p><p>The minibatch size <code>N</code> is extracted from `sparse_shape[0]`.</p></div></div><divclass="top"><pclass="src"><aname="v:initializeTableFromTextFile"class="def">initializeTableFromTextFile</a><ahref="src/TensorFlow-GenOps-Core.html#initializeTableFromTextFile"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>key_index</strong>: Column index in a line to get the table <code>key</code> values from.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>value_index</strong>: Column index that represents information of a line to get the table
<code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code> values from.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>table_handle</strong>: Handle to a table which will be initialized.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>filename</strong>: Filename of a vocabulary text file.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Initializes a table from a text file.</p><p>It inserts one key-value pair into the table for each line of the file.
The key and value is extracted from the whole line content, elements from the
split line based on <code>delimiter</code> or the line number (starting from zero).
Where to extract the key and value from a line is specified by <code>key_index</code> and
<code>value_index</code>.</p><ul><li>A value of -1 means use the line number(starting from zero), expects <code>int64</code>.</li><li>A value of -2 means use the whole line content, expects <code>string</code>.</li><li>A value >= 0 means use the index (starting at zero) of the split line based
on <code>delimiter</code>.</li></ul></div></div><divclass="top"><pclass="src"><aname="v:decodePng"class="def">decodePng</a><ahref="src/TensorFlow-GenOps-Core.html#decodePng"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` dtype)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>contents</strong>: 0-D. The PNG-encoded image.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype</td><tdclass="doc"><p><strong>image</strong>: 3-D with shape `[height, width, channels]`.</p></td></tr></table></div><divclass="doc"><p>Decode a PNG-encoded image to a uint8 or uint16 tensor.</p><p>The attr <code>channels</code> indicates the desired number of color channels for the
decoded image.</p><p>Accepted values are:</p><ul><li>0: Use the number of channels in the PNG-encoded image.</li><li>1: output a grayscale image.</li><li>3: output an RGB image.</li><li>4: output an RGBA image.</li></ul><p>If needed, the PNG-encoded image is transformed to match the requested number
of color channels.</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArraySizeV2"class="def">tensorArraySizeV2</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArraySizeV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a TensorArray (output of TensorArray or TensorArrayGrad).</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong>: A float scalar that enforces proper chaining of operations.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>size</strong>: The current size of the TensorArray.</p></td></tr></table></div><divclass="doc"><p>Get the current size of the TensorArray.</p></div></div><divclass="top"><pclass="src"><aname="v:div"class="def">div</a><ahref="src/TensorFlow-GenOps-Core.html#div"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns x / y element-wise.</p><ul><li>NOTE*: <code>Div</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:logUniformCandidateSampler"class="def">logUniformCandidateSampler</a><ahref="src/TensorFlow-GenOps-Core.html#logUniformCandidateSampler"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_sampled</strong>: Number of candidates to randomly sample per batch.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_true</strong>: Number of true labels per context.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>range_max</strong>: The sampler will sample integers from the interval [0, range_max).</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>unique</strong>: If unique is true, we sample with rejection, so that all sampled
candidates in a batch are unique. This requires some approximation to
estimate the post-rejection sampling probabilities.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>true_classes</strong>: A batch_size * num_true matrix, in which each row contains the
IDs of the num_true target_classes in the corresponding original label.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>sampled_candidates</strong>, <strong>true_expected_count</strong>, <strong>sampled_expected_count</strong>)</p><ul><li><strong>sampled_candidates</strong>: A vector of length num_sampled, in which each element is
the ID of a sampled candidate.</li><li><strong>true_expected_count</strong>: A batch_size * num_true matrix, representing
the number of times each candidate is expected to occur in a batch
of sampled candidates. If unique=true, then this is a probability.</li><li><strong>sampled_expected_count</strong>: A vector of length num_sampled, for each sampled
candidate representing the number of times the candidate is expected
to occur in a batch of sampled candidates. If unique=true, then this is a
probability.</li></ul></td></tr></table></div><divclass="doc"><p>Generates labels for candidate sampling with a log-uniform distribution.</p><p>See explanations of candidate sampling and the data formats at
go/candidate-sampling.</p><p>For each batch, this op picks a single set of sampled candidate labels.</p><p>The advantages of sampling candidates per-batch are simplicity and the
possibility of efficient dense matrix multiplication. The disadvantage is that
the sampled candidates must be chosen independently of the context and of the
true labels.</p></div></div><divclass="top"><pclass="src"><aname="v:barrier"class="def">barrier</a><ahref="src/TensorFlow-GenOps-Core.html#barrier"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>)</td><tdclass="doc"><p><strong>handle</strong>: The handle to the barrier.</p></td></tr></table></div><divclass="doc"><p>Defines a barrier that persists across different graph executions.</p><p>A barrier represents a key-value map, where each key is a string, and
each value is a tuple of tensors.</p><p>At runtime, the barrier contains <code>complete</code> and <code>incomplete</code>
elements. A complete element has defined tensors for all components of
its value tuple, and may be accessed using BarrierTakeMany. An
incomplete element has some undefined components in its value tuple,
and may be updated using BarrierInsertMany.</p></div></div><divclass="top"><pclass="src"><aname="v:createVariableOp"class="def">createVariableOp</a><ahref="src/TensorFlow-GenOps-Core.html#createVariableOp"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ResourceHandle">ResourceHandle</a> dtype</td><tdclass="doc"><p><strong>resource</strong>: handle to the resource in which to store the variable.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 dtype</td><tdclass="doc"><p><strong>value</strong>: the value to set the new tensor to use.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Creates a variable resource.</p></div></div><divclass="top"><pclass="src"><aname="v:accumulatorApplyGradient"class="def">accumulatorApplyGradient</a><ahref="src/TensorFlow-GenOps-Core.html#accumulatorApplyGradient"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` dtype)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a accumulator.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>local_step</strong>: The local_step value at which the gradient was computed.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 dtype</td><tdclass="doc"><p><strong>gradient</strong>: A tensor of the gradient to be accumulated.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Applies a gradient to a given accumulator. Does not add if local_step is lesser</p><p>than the accumulator's global_step.</p></div></div><divclass="top"><pclass="src"><aname="v:randomStandardNormal"class="def">randomStandardNormal</a><ahref="src/TensorFlow-GenOps-Core.html#randomStandardNormal"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorF
for each batch.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype)</td><tdclass="doc"><p><strong>output</strong>: A matrix of shape num_batches x samples_per_batch, filled with random
truncated normal values using the parameters for each row.</p></td></tr></table></div><divclass="doc"><p>Outputs random values from a normal distribution. The parameters may each be a</p><p>scalar which applies to the entire output, or a vector of length shape[0] which
stores the parameters for each batch.</p></div></div><divclass="top"><pclass="src"><aname="v:accumulatorSetGlobalStep"class="def">accumulatorSetGlobalStep</a><ahref="src/TensorFlow-GenOps-Core.html#accumulatorSetGlobalStep"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to an accumulator.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>new_global_step</strong>: The new global_step value to set.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Updates the accumulator with a new value for global_step. Logs warning if the</p><p>accumulator's value is already higher than new_global_step.</p></div></div><divclass="top"><pclass="src"><aname="v:resizeBilinear"class="def">resizeBilinear</a><ahref="src/TensorFlow-GenOps-Core.html#resizeBilinear"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>images</strong>: 4-D with shape `[batch, height, width, channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>size</strong>: = A 1-D int32 Tensor of 2 elements: `new_height, new_width`. The
new size for the images.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>resized_images</strong>: 4-D with shape
`[batch, new_height, new_width, channels]`.</p></td></tr></table></div><divclass="doc"><p>Resize <code>images</code> to <code><ahref="TensorFlow-GenOps-Core.html#v:size">size</a></code> using bilinear interpolation.</p><p>Input images can be of different types but output images are always float.</p></div></div><divclass="top"><pclass="src"><aname="v:quantizeV2"class="def">quantizeV2</a><ahref="src/TensorFlow-GenOps-Core.html#quantizeV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>min_range</strong>: The minimum scalar value possibly produced for the input.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>max_range</strong>: The maximum scalar value possibly produced for the input.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>output</strong>, <strong>output_min</strong>, <strong>output_max</strong>)</p><ul><li><strong>output</strong>: The quantized data produced from the float input.</li><li><strong>output_min</strong>: The actual minimum scalar value used for the output.</li><li><strong>output_max</strong>: The actual maximum scalar value used for the output.</li></ul></td></tr></table></div><divclass="doc"><p>Quantize the <code>input</code> tensor of type float to <code>output</code> tensor of type <code>T</code>.</p><dl><dt>min_range, max_range</dt><dd>are scalar floats that specify the range for
the <code>input</code> data. The <code>mode</code> attribute controls exactly which calculations are
used to convert the float values to their quantized equivalents.</dd></dl><p>In <code>MIN_COMBINED</code> mode, each value of the tensor will undergo the following:</p><p>```
here `range(T) = numeric_limits<ahref="T">T</a>::max() - numeric_limits<ahref="T">T</a>::min()`</p><ul><li>MIN_COMBINED Mode Example*</li></ul><p>Assume the input is type float and has a possible range of [0.0, 6.0] and the
output type is quint8 ([0, 255]). The min_range and max_range values should be
specified as 0.0 and 6.0. Quantizing from float to quint8 will multiply each
value of the input by 255/6 and cast to quint8.</p><p>If the output type was qint8 ([-128, 127]), the operation will additionally
subtract each value by 128 prior to casting, so that the range of values aligns
with the range of qint8.</p><p>If the mode is <code>MIN_FIRST</code>, then this approach is used:</p><p>```
```</p><p>The biggest difference between this and MIN_COMBINED is that the minimum range
is rounded first, before it's subtracted from the rounded value. With
MIN_COMBINED, a small bias is introduced where repeated iterations of quantizing
and dequantizing will introduce a larger and larger error.</p><p>One thing to watch out for is that the operator may choose to adjust the
requested minimum and maximum values slightly during the quantization process,
so you should always use the output ports as the range for further calculations.
For example, if the requested minimum and maximum values are close to equal,
they will be separated by a small epsilon value to prevent ill-formed quantized
buffers from being created. Otherwise, you can end up with buffers where all the
quantized values map to the same float value, which causes problems for
operations that have to perform further calculations on them.</p></div></div><divclass="top"><pclass="src"><aname="v:decodeJpeg"class="def">decodeJpeg</a><ahref="src/TensorFlow-GenOps-Core.html#decodeJpeg"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>contents</strong>: 0-D. The JPEG-encoded image.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a></td><tdclass="doc"><p><strong>image</strong>: 3-D with shape `[height, width, channels]`..</p></td></tr></table></div><divclass="doc"><p>Decode a JPEG-encoded image to a uint8 tensor.</p><p>The attr <code>channels</code> indicates the desired number of color channels for the
decoded image.</p><p>Accepted values are:</p><ul><li>0: Use the number of channels in the JPEG-encoded image.</li><li>1: output a grayscale image.</li><li>3: output an RGB image.</li></ul><p>If needed, the JPEG-encoded image is transformed to match the requested number
of color channels.</p><p>The attr <code>ratio</code> allows downscaling the image by an integer factor during
decoding. Allowed values are: 1, 2, 4, and 8. This is much faster than
downscaling the image later.</p></div></div><divclass="top"><pclass="src"><aname="v:pow"class="def">pow</a><ahref="src/TensorFlow-GenOps-Core.html#pow"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Computes the power of one value to another.</p><p>Given a tensor <code>x</code> and a tensor <code>y</code>, this operation computes \(x^y\) for
corresponding elements in <code>x</code> and <code>y</code>. For example:</p><p>```
# tensor <code>x</code> is [[2, 2]], [3, 3]]
# tensor <code>y</code> is [[8, 16], [2, 3]]
tf.pow(x, y) ==> [[256, 65536], [9, 27]]
```</p></div></div><divclass="top"><pclass="src"><aname="v:loopCond"class="def">loopCond</a><ahref="src/TensorFlow-GenOps-Core.html#loopCond"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>input</strong>: A boolean scalar, representing the branch predicate of the Switch op.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>output</strong>: The same tensor as <code>input</code>.</p></td></tr></table></div><divclass="doc"><p>Forwards the input to the output.</p><p>This operator represents the loop termination condition used by the
"pivot" switches of a loop.</p></div></div><divclass="top"><pclass="src"><aname="v:readFile"class="def">readFile</a><ahref="src/TensorFlow-GenOps-Core.html#readFile"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>filename</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>contents</strong></p></td></tr></table></div><divclass="doc"><p>Reads and outputs the entire contents of the input filename.</p></div></div><divclass="top"><pclass="src"><aname="v:imag"class="def">imag</a><ahref="src/TensorFlow-GenOps-Core.html#imag"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tout, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` tout)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> tout</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Returns the imaginary part of a complex number.</p><p>Given a tensor <code>input</code> of complex numbers, this operation returns a tensor of
type <code>float</code> that is the imaginary part of each element in <code>input</code>. All
elements in <code>input</code> must be complex numbers of the form \(a + bj\), where *a*
is the real part and *b* is the imaginary part returned by this operation.</p><p>For example:</p><p>```
# tensor <code>input</code> is [-2.25 + 4.75j, 3.25 + 5.75j]
tf.imag(input) ==> [4.75, 5.75]
```</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArrayGrad"class="def">tensorArrayGrad</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArrayGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>)</td><tdclass="doc"><p><strong>grad_handle</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:histogramSummary"class="def">histogramSummary</a><ahref="src/TensorFlow-GenOps-Core.html#histogramSummary"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>tag</strong>: Scalar. Tag to use for the <code><ahref="Summary.html#v:Value">Value</a></code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>values</strong>: Any shape. Values to use to build the histogram.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>summary</strong>: Scalar. Serialized <code>Summary</code> protocol buffer.</p></td></tr></table></div><divclass="doc"><p>Outputs a <code>Summary</code> protocol buffer with a histogram.</p><p>The generated
has one summary value containing a histogram for <code>values</code>.</p><p>This op reports an <code>InvalidArgument</code> error if any value is not finite.</p></div></div><divclass="top"><pclass="src"><aname="v:conv3DBackpropInputV2"class="def">conv3DBackpropInputV2</a><ahref="src/TensorFlow-GenOps-Core.html#conv3DBackpropInputV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>input_sizes</strong>: An integer vector representing the tensor shape of <code>input</code>,
<code>in_channels</code> must match between <code>input</code> and <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>out_backprop</strong>: Backprop signal of shape `[batch, out_depth, out_rows, out_cols,
out_channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Computes the gradients of 3-D convolution with respect to the input.</p></div></div><divclass="top"><pclass="src"><aname="v:resizeBilinearGrad"class="def">resizeBilinearGrad</a><ahref="src/TensorFlow-GenOps-Core.html#resizeBilinearGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>grads</strong>: 4-D with shape `[batch, height, width, channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>original_image</strong>: 4-D with shape `[batch, orig_height, orig_width, channels]`,
The image tensor that was resized.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 4-D with shape `[batch, orig_height, orig_width, channels]`.
Gradients with respect to the input image. Input image must have been
float or double.</p></td></tr></table></div><divclass="doc"><p>Computes the gradient of bilinear interpolation.</p></div></div><divclass="top"><pclass="src"><aname="v:addManySparseToTensorsMap"class="def">addManySparseToTensorsMap</a><ahref="src/TensorFlow-GenOps-Core.html#addManySparseToTensorsMap"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sparse_indices</strong>: 2-D. The <code>indices</code> of the minibatch <code>SparseTensor</code>.
`sparse_indices[:, 0]` must be ordered values in `[0, N)`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>sparse_values</strong>: 1-D. The <code>values</code> of the minibatch <code>SparseTensor</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sparse_shape</strong>: 1-D. The <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> of the minibatch <code>SparseTensor</code>.
The minibatch size `N == sparse_shape[0]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p><strong>sparse_handles</strong>: 1-D. The handles of the <code>SparseTensor</code> now stored in the
<code>SparseTensorsMap</code>. Shape: `[N]`.</p></td></tr></table></div><divclass="doc"><p>Add an <code>N</code>-minibatch <code>SparseTensor</code> to a <code>SparseTensorsMap</code>, return <code>N</code> handles.</p><p>A <code>SparseTensor</code> of rank <code>R</code> is represented by three tensors: <code>sparse_indices</code>,
<code>sparse_values</code>, and <code>sparse_shape</code>, where</p><p>```sparse_indices.shape[1] == sparse_shape.shape[0] == R```</p><p>An <code>N</code>-minibatch of <code>SparseTensor</code> objects is represented as a <code>SparseTensor</code>
having a first <code>sparse_indices</code> column taking values between `[0, N)`, where
the minibatch size `N == sparse_shape[0]`.</p><p>The input <code>SparseTensor</code> must have rank <code>R</code> greater than 1, and the first
dimension is treated as the minibatch dimension. Elements of the <code>SparseTensor</code>
must be sorted in increasing order of this first dimension. The stored
<code>SparseTensor</code> objects pointed to by each row of the output <code>sparse_handles</code>
will have rank `R-1`.</p><p>The <code>SparseTensor</code> values can then be read out as part of a minibatch by passing
the given keys as vector elements to <code>TakeManySparseFromTensorsMap</code>. To ensure
the correct <code>SparseTensorsMap</code> is accessed, ensure that the same
<code>container</code> and <code>shared_name</code> are passed to that Op. If no <code>shared_name</code>
is provided here, instead use the *name* of the Operation created by calling
<code>AddManySparseToTensorsMap</code> as the <code>shared_name</code> passed to
<code>TakeManySparseFromTensorsMap</code>. Ensure the Operations are colocated.</p></div></div><divclass="top"><pclass="src"><aname="v:batchIFFT"class="def">batchIFFT</a><ahref="src/TensorFlow-GenOps-Core.html#batchIFFT"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:batchMatrixDeterminant"class="def">batchMatrixDeterminant</a><ahref="src/TensorFlow-GenOps-Core.html#batchMatrixDeterminant"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:deleteSessionTensor"class="def">deleteSessionTensor</a><ahref="src/TensorFlow-GenOps-Core.html#deleteSessionTensor"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle for a tensor stored in the session state.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Delete the tensor specified by its handle in the session.</p></div></div><divclass="top"><pclass="src"><aname="v:lookupTableSize"class="def">lookupTableSize</a><ahref="src/TensorFlow-GenOps-Core.html#lookupTableSize"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>table_handle</strong>: Handle to the table.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p><strong>size</strong>: Scalar that contains number of elements in the table.</p></td></tr></table></div><divclass="doc"><p>Computes the number of elements in the given table.</p></div></div><divclass="top"><pclass="src"><aname="v:relu"class="def">relu</a><ahref="src/TensorFlow-GenOps-Core.html#relu"class="link">Source</a></p><divclass="subs
```</p><p>Each `data[i].shape` must start with the corresponding `indices[i].shape`,
and the rest of `data[i].shape` must be constant w.r.t. <code>i</code>. That is, we
must have `data[i].shape = indices[i].shape + constant`. In terms of this
<code>constant</code>, the output shape is</p><p>merged.shape = [max(indices)] + constant</p><p>Values are merged in order, so if an index appears in both `indices[m][i]` and
`indices[n][j]` for `(m,i) < (n,j)` the slice `data[n][j]` will appear in the
<ahref="/div">/div</a></p></div></div><divclass="top"><pclass="src"><aname="v:lookupTableFind"class="def">lookupTableFind</a><ahref="src/TensorFlow-GenOps-Core.html#lookupTableFind"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tin, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tout)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>table_handle</strong>: Handle to the table.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tin</td><tdclass="doc"><p><strong>keys</strong>: Any shape. Keys to look up.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 tout</td><tdclass="doc"><p><strong>default_value</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> tout)</td><tdclass="doc"><p><strong>values</strong>: Same shape as <code>keys</code>. Values found in the table, or <code>default_values</code>
for missing keys.</p></td></tr></table></div><divclass="doc"><p>Looks up keys in a table, outputs the corresponding values.</p><p>The tensor <code>keys</code> must of the same type as the keys of the table.
The output <code>values</code> is of the type of the table values.</p><p>The scalar <code>default_value</code> is the value output for keys not present in the
table. It must also be of the same type as the table values.</p></div></div><divclass="top"><pclass="src"><aname="v:sampleDistortedBoundingBox"class="def">sampleDistortedBoundingBox</a><ahref="src/TensorFlow-GenOps-Core.html#sampleDistortedBoundingBox"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>image_size</strong>: 1-D, containing `[height, width, channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>bounding_boxes</strong>: 3-D with shape `[batch, N, 4]` describing the N bounding boxes
associated with the image.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>begin</strong>, <strong>size</strong>, <strong>bboxes</strong>)</p><ul><li><strong>begin</strong>: 1-D, containing `[offset_height, offset_width, 0]`. Provide as input to
`tf.slice`.</li><li><strong>size</strong>: 1-D, containing `[target_height, target_width, -1]`. Provide as input to
`tf.slice`.</li><li><strong>bboxes</strong>: 3-D with shape `[1, 1, 4]` containing the distorted bounding box.
Provide as input to `tf.image.draw_bounding_boxes`.</li></ul></td></tr></table></div><divclass="doc"><p>Generate a single randomly distorted bounding box for an image.</p><p>Bounding box annotations are often supplied in addition to ground-truth labels
in image recognition or object localization tasks. A common technique for
training such a system is to randomly distort an image while preserving
its content, i.e. *data augmentation*. This Op outputs a randomly distorted
localization of an object, i.e. bounding box, given an <code>image_size</code>,
<code>bounding_boxes</code> and a series of constraints.</p><p>The output of this Op is a single bounding box that may be used to crop the
original image. The output is returned as 3 tensors: <code>begin</code>, <code><ahref="TensorFlow-GenOps-Core.html#v:size">size</a></code> and
<code>bboxes</code>. The first 2 tensors can be fed directly into `tf.slice` to crop the
image. The latter may be supplied to `tf.image.draw_bounding_boxes` to visualize
what the bounding box looks like.</p><p>Bounding boxes are supplied and returned as `[y_min, x_min, y_max, x_max]`. The
bounding box coordinates are floats in `[0.0, 1.0]` relative to the width and
height of the underlying image.</p><p>For example,</p><p>```python
tf.image_summary(<code>images_with_box</code>, image_with_box)</p><p># Employ the bounding box to distort the image.
distorted_image = tf.slice(image, begin, size)
```</p><p>Note that if no bounding box information is available, setting
`use_image_if_no_bounding_boxes = true` will assume there is a single implicit
bounding box covering the whole image. If <code>use_image_if_no_bounding_boxes</code> is
false and no bounding boxes are supplied, an error is raised.</p></div></div><divclass="top"><pclass="src"><aname="v:splitV"class="def">splitV</a><ahref="src/TensorFlow-GenOps-Core.html#splitV"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tlen, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tlen)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_split</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>value</strong>: The tensor to split.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tlen</td><tdclass="doc"><p><strong>size_splits</strong>: list containing the sizes of each output tensor along the split
dimension. Must sum to the dimension of value along split_dim.
Can contain one -1 indicating that dimension is to be inferred.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>split_dim</strong>: 0-D. The dimension along which to split. Must be in the range
`[0, rank(value))`.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t]</td><tdclass="doc"><p><strong>output</strong>: Tensors whose shape matches that of <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>
except along <code>split_dim</code>, where their sizes are
`size_splits[i]`.</p></td></tr></table></div><divclass="doc"><p>Splits a tensor into <code>num_split</code> tensors along one dimension.</p></div></div><divclass="top"><pclass="src"><aname="v:fusedPadConv2D"class="def">fusedPadConv2D</a><ahref="src/TensorFlow-GenOps-Core.html#fusedPadConv2D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: 4-D with shape `[batch, in_height, in_width, in_channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>paddings</strong>: A two-column matrix specifying the padding sizes. The number of
rows must be the same as the rank of <code>input</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>filter</strong>: 4-D with shape
`[filter_height, filter_width, in_channels, out_channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Performs a padding as a preprocess during a convolution.</p><p>Similar to FusedResizeAndPadConv2d, this op allows for an optimized
implementation where the spatial padding transformation stage is fused with the
im2col lookup, but in this case without the bilinear filtering required for
resizing. Fusing the padding prevents the need to write out the intermediate
results as whole tensors, reducing memory pressure, and we can get some latency
gains by merging the transformation calculations.
The data_format attribute for Conv2D isn't supported by this op, and <code>NHWC</code>
order is used instead.
Internally this op uses a single per-graph scratch buffer, which means that it
will block if multiple versions are being run in parallel. This is because this
operator is primarily an optimization to minimize memory usage.</p></div></div><divclass="top"><pclass="src"><aname="v:barrierInsertMany"class="def">barrierInsertMany</a><ahref="src/TensorFlow-GenOps-Core.html#barrierInsertMany"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>component_index</strong>: The component of the barrier elements that is being assigned.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a barrier.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>keys</strong>: A one-dimensional tensor of keys, with length n.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>values</strong>: An any-dimensional tensor of values, which are associated with the
respective keys. The 0th dimension must have length n.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>For each key, assigns the respective value to the specified component.</p><p>If a key is not found in the barrier, this operation will create a new
incomplete element. If a key is found in the barrier, and the element
already has a value at component_index, this operation will fail with
INVALID_ARGUMENT, and leave the barrier in an undefined state.</p></div></div><divclass="top"><pclass="src"><aname="v:abort"class="def">abort</a> :: <ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a><ahref="src/TensorFlow-GenOps-Core.html#abort"class="link">Source</a></p><divclass="doc"><p>Raise a exception to abort the process when called.</p><p>Returns nothing but an exception.</p></div></div><divclass="top"><pclass="src"><aname="v:maxPoolWithArgmax"class="def">maxPoolWithArgmax</a><ahref="src/TensorFlow-GenOps-Core.html#maxPoolWithArgmax"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> targmax, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` targmax, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: 4-D with shape `[batch, height, width, channels]`. Input to pool over.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> targmax)</td><tdclass="doc"><p>(<strong>output</strong>, <strong>argmax</strong>)</p><ul><li><strong>output</strong>: The max pooled output tensor.</li><li><strong>argmax</strong>: 4-D. The flattened indices of the max values chosen for each output.</li></ul></td></tr></table></div><divclass="doc"><p>Performs max pooling on the input and outputs both max values and indices.</p><p>The indices in <code>argmax</code> are flattened, so that a maximum value at position
`[b, y, x, c]` becomes flattened index
`((b * height + y) * width + x) * channels + c`.</p></div></div><divclass="top"><pclass="src"><aname="v:refEnter"class="def">refEnter</a><ahref="src/TensorFlow-GenOps-Core.html#refEnter"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>data</strong>: The tensor to be made available to the child frame.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output</strong>: The same tensor as `data`.</p></td></tr></table></div><divclass="doc"><p>Creates or finds a child frame, and makes `data` available to the child frame.</p><p>The unique <code>frame_name</code> is used by the <code>Executor</code> to identify frames. If
<code>is_constant</code> is true, <code>output</code> is a constant in the child frame; otherwise
it may be changed in the child frame. At most <code>parallel_iterations</code> iterations
are run in parallel in the child frame.</p></div></div><divclass="top"><pclass="src"><aname="v:dequantize"class="def">dequantize</a><ahref="src/TensorFlow-GenOps-Core.html#dequantize"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>min_range</strong>: The minimum scalar value possibly produced for the input.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>max_range</strong>: The maximum scalar value possibly produced for the input.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Dequantize the <code>input</code> tensor into a float Tensor.</p><dl><dt>min_range, max_range</dt><dd>are scalar floats that specify the range for
the <code>input</code> data. The <code>mode</code> attribute controls exactly which calculations are
used to convert the float values to their quantized equivalents.</dd></dl><p>In <code>MIN_COMBINED</code> mode, each value of the tensor will undergo the following:</p><p>```
here `range(T) = numeric_limits<ahref="T">T</a>::max() - numeric_limits<ahref="T">T</a>::min()`</p><ul><li>MIN_COMBINED Mode Example*</li></ul><p>If the input comes from a QuantizedRelu6, the output type is
quint8 (range of 0-255) but the possible range of QuantizedRelu6 is
0-6. The min_range and max_range values are therefore 0.0 and 6.0.
Dequantize on quint8 will take each value, cast to float, and multiply
by 6 / 255.
Note that if quantizedtype is qint8, the operation will additionally add
each value by 128 prior to casting.</p><p>If the mode is <code>MIN_FIRST</code>, then this approach is used:</p><p>```
result = range_min + ((input - numeric_limits<ahref="T">T</a>::min()) * range_scale)
```</p></div></div><divclass="top"><pclass="src"><aname="v:drawBoundingBoxes"class="def">drawBoundingBoxes</a><ahref="src/TensorFlow-GenOps-Core.html#drawBoundingBoxes"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>images</strong>: 4-D with shape `[batch, height, width, depth]`. A batch of images.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>boxes</strong>: 3-D with shape `[batch, num_bounding_boxes, 4]` containing bounding
boxes.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 4-D with the same shape as <code>images</code>. The batch of input images with
bounding boxes drawn on the images.</p></td></tr></table></div><divclass="doc"><p>Draw bounding boxes on a batch of images.</p><p>Outputs a copy of <code>images</code> but draws on top of the pixels zero or more bounding
boxes specified by the locations in <code>boxes</code>. The coordinates of the each
bounding box in <code>boxes</code> are encoded as `[y_min, x_min, y_max, x_max]`. The
bounding box coordinates are floats in `[0.0, 1.0]` relative to the width and
height of the underlying image.</p><p>For example, if an image is 100 x 200 pixels and the bounding box is
`[0.1, 0.2, 0.5, 0.9]`, the bottom-left and upper-right coordinates of the
bounding box will be `(10, 40)` to `(50, 180)`.</p><p>Parts of the bounding box may fall outside the image.</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArraySplit"class="def">tensorArraySplit</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArraySplit"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>value</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>lengths</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>flow_out</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:stringToHashBucketFast"class="def">stringToHashBucketFast</a><ahref="src/TensorFlow-GenOps-Core.html#stringToHashBucketFast"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_buckets</strong>: The number of buckets.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>input</strong>: The strings to assign a hash bucket.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>output</strong>: A Tensor of the same shape as the input <code>string_tensor</code>.</p></td></tr></table></div><divclass="doc"><p>Converts each string in the input Tensor to its hash mod by a number of buckets.</p><p>The hash function is deterministic on the content of the string within the
process and will never change. However, it is not suitable for cryptography.
This function may be used when CPU time is scarce and inputs are trusted or
unimportant. There is a risk of adversaries constructing inputs that all hash
to the same bucket. To prevent this problem, use a strong hash function with
`tf.string_to_hash_bucket_strong`.</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArrayScatter"class="def">tensorArrayScatter</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArrayScatter"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>indices</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>value</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>flow_out</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:oneHot"class="def">oneHot</a><ahref="src/TensorFlow-GenOps-Core.html#oneHot"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tI, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` tI)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 tI</td><tdclass="doc"><p><strong>indices</strong>: A tensor of indices.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>depth</strong>: A scalar defining the depth of the one hot dimension.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>on_value</strong>: A scalar defining the value to fill in output when `indices[j] = i`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>off_value</strong>: A scalar defining the value to fill in output when `indices[j] != i`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The one-hot tensor.</p></td></tr></table></div><divclass="doc"><p>Returns a one-hot tensor.</p><p>The locations represented by indices in <code>indices</code> take value <code>on_value</code>,
while all other locations take value <code>off_value</code>.</p><p>If the input <code>indices</code> is rank <code>N</code>, the output will have rank `N+1`,
The new axis is created at dimension <code>axis</code> (default: the new axis is
appended at the end).</p><p>If <code>indices</code> is a scalar the output shape will be a vector of length <code>depth</code>.</p><p>If <code>indices</code> is a vector of length <code>features</code>, the output shape will be:
```
features x depth if axis == -1
depth x features if axis == 0
```</p><p>If <code>indices</code> is a matrix (batch) with shape `[batch, features]`,
the output shape will be:
```
batch x features x depth if axis == -1
batch x depth x features if axis == 1
depth x batch x features if axis == 0
```</p><p>Examples
=========</p><p>Suppose that</p><p>```
indices = [0, 2, -1, 1]
depth = 3
on_value = 5.0
off_value = 0.0
axis = -1
```</p><p>Then output is `[4 x 3]`:</p><p>```output =
[5.0 0.0 0.0] // one_hot(0)
[0.0 0.0 5.0] // one_hot(2)
[0.0 0.0 0.0] // one_hot(-1)
[0.0 5.0 0.0] // one_hot(1)
```</p><p>Suppose that</p><p>```
indices = [0, 2, -1, 1]
depth = 3
on_value = 0.0
off_value = 3.0
axis = 0
```</p><p>Then output is `[3 x 4]`:</p><p>```output =
[0.0 3.0 3.0 3.0]
[3.0 3.0 3.0 0.0]
[3.0 3.0 3.0 3.0]
[3.0 0.0 3.0 3.0]
// ^ one_hot(0)
// ^ one_hot(2)
// ^ one_hot(-1)
// ^ one_hot(1)
```
Suppose that</p><p>```
indices = [[0, 2], [1, -1]]
depth = 3
on_value = 1.0
off_value = 0.0
axis = -1
```</p><p>Then output is `[2 x 2 x 3]`:</p><p>```output =
]```</p></div></div><divclass="top"><pclass="src"><aname="v:batchIFFT3D"class="def">batchIFFT3D</a><ahref="src/TensorFlow-GenOps-Core.html#batchIFFT3D"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> (<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:decodeRaw"class="def">decodeRaw</a><ahref="src/TensorFlow-GenOps-Core.html#decodeRaw"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` out_type)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>bytes</strong>: All the elements must have the same length.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_type</td><tdclass="doc"><p><strong>output</strong>: A Tensor with one more dimension than the input <code>bytes</code>. The
added dimension will have size equal to the length of the elements
of <code>bytes</code> divided by the number of bytes to represent <code>out_type</code>.</p></td></tr></table></div><divclass="doc"><p>Reinterpret the bytes of a string as a vector of numbers.</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArrayPack"class="def">tensorArrayPack</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArrayPack"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype)</td><tdclass="doc"><p><strong>value</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:applyProximalAdagrad"class="def">applyProximalAdagrad</a><ahref="src/TensorFlow-GenOps-Core.html#applyProximalAdagrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>accum</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>lr</strong>: Scaling factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>l1</strong>: L1 regularization. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>l2</strong>: L2 regularization. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../te
prox_v = var - lr * grad * (1 / sqrt(accum))
var = sign(prox_v)/(1+lr*l2) * max{|prox_v|-lr*l1,0}</p></div></div><divclass="top"><pclass="src"><aname="v:sparseAccumulatorApplyGradient"class="def">sparseAccumulatorApplyGradient</a><ahref="src/TensorFlow-GenOps-Core.html#sparseAccumulatorApplyGradient"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` dtype)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>has_known_shape</strong>: Boolean indicating whether gradient_shape is unknown, in which
case the input is ignored during validation.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a accumulator.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>local_step</strong>: The local_step value at which the sparse gradient was computed.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>gradient_indices</strong>: Indices of the sparse gradient to be accumulated. Must be a
vector.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 dtype</td><tdclass="doc"><p><strong>gradient_values</strong>: Values are the non-zero slices of the gradient, and must have
the same first dimension as indices, i.e., the nnz represented by indices and
values must be consistent.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>gradient_shape</strong>: Shape of the sparse gradient to be accumulated.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Applies a sparse gradient to a given accumulator. Does not add if local_step is</p><p>lesser than the accumulator's global_step.</p></div></div><divclass="top"><pclass="src"><aname="v:add"class="def">add</a><ahref="src/TensorFlow-GenOps-Core.html#add"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns x + y element-wise.</p><ul><li>NOTE*: <code>Add</code> supports broadcasting. <code>AddN</code> does not. More about broadcasting
A tensor of indices into ref.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>updates</strong>: A Tensor. Must have the same type as ref. A tensor of updated values
to subtract from ref.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output_ref</strong>: Same as ref. Returned as a convenience for operations that want
to use the updated values after the update is done.</p></td></tr></table></div><divclass="doc"><p>Applies sparse subtraction between <code>updates</code> and individual values or slices</p><p>within a given variable according to <code>indices</code>.</p><p><code>ref</code> is a <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> with rank <code>P</code> and <code>indices</code> is a <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> of rank <code>Q</code>.</p><p><code>indices</code> must be integer tensor, containing indices into <code>ref</code>.
It must be shape `[d_0, ..., d_{Q-2}, K]` where `0 < K <= P`.</p><p>The innermost dimension of <code>indices</code> (with length <code>K</code>) corresponds to
indices into elements (if `K = P`) or slices (if `K < P`) along the <code>K</code>th
dimension of <code>ref</code>.</p><p><code>updates</code> is <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> of rank `Q-1+P-K` with shape:</p><p>```
```</p><p>For example, say we want to subtract 4 scattered elements from a rank-1 tensor
with 8 elements. In Python, that subtraction would look like this:</p><p>ref = tf.Variable([1, 2, 3, 4, 5, 6, 7, 8])
indices = tf.constant([[4], [3], [1], [7]])
updates = tf.constant([9, 10, 11, 12])
sub = tf.scatter_nd_sub(ref, indices, updates)
with tf.Session() as sess:
print sess.run(sub)</p><p>The resulting update to ref would look like this:</p><dl><dt>1, -9, 3, -6, -4, 6, 7, -4</dt><dd></dd></dl><p>See <ahref="#scatter_nd">tf.scatter_nd</a> for more details about how to make updates to
slices.</p></div></div><divclass="top"><pclass="src"><aname="v:restoreSlice"class="def">restoreSlice</a><ahref="src/TensorFlow-GenOps-Core.html#restoreSlice"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dt</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>file_pattern</strong>: Must have a single element. The pattern of the files from
which we read the tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>tensor_name</strong>: Must have a single element. The name of the tensor to be
restored.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>shape_and_slice</strong>: Scalar. The shapes and slice specifications to use when
restoring a tensors.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dt</td><tdclass="doc"><p><strong>tensor</strong>: The restored tensor.</p></td></tr></table></div><divclass="doc"><p>Restores a tensor from checkpoint files.</p><p>This is like <code>Restore</code> except that restored tensor can be listed as filling
only a slice of a larger tensor. <code>shape_and_slice</code> specifies the shape of the
larger tensor and the slice that the restored tensor covers.</p><p>The <code>shape_and_slice</code> input has the same format as the
elements of the <code>shapes_and_slices</code> input of the <code>SaveSlices</code> op.</p></div></div><divclass="top"><pclass="src"><aname="v:assignAdd"class="def">assignAdd</a><ahref="src/TensorFlow-GenOps-Core.html#assignAdd"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ref</strong>: Should be from a <code>Variable</code> node.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>value</strong>: The value to be added to the variable.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output_ref</strong>: = Same as "ref". Returned as a convenience for operations that want
to use the new value after the variable has been updated.</p></td></tr></table></div><divclass="doc"><p>Update <code>ref</code> by adding <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code> to it.</p><p>This operation outputs "ref" after the update is done.
This makes it easier to chain operations that need to use the reset value.</p></div></div><divclass="top"><pclass="src"><aname="v:greater"class="def">greater</a><ahref="src/TensorFlow-GenOps-Core.html#greater"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns the truth value of (x > y) element-wise.</p><ul><li>NOTE*: <code>Greater</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:readerNumWorkUnitsCompleted"class="def">readerNumWorkUnitsCompleted</a><ahref="src/TensorFlow-GenOps-Core.html#readerNumWorkUnitsCompleted"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>reader_handle</strong>: Handle to a Reader.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p><strong>units_completed</strong></p></td></tr></table></div><divclass="doc"><p>Returns the number of work units this Reader has finished processing.</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArrayGatherV2"class="def">tensorArrayGatherV2</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArrayGatherV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a TensorArray.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>indices</strong>: The locations in the TensorArray from which to read tensor elements.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong>: A float scalar that enforces proper chaining of operations.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype</td><tdclass="doc"><p><strong>value</strong>: All of the elements in the TensorArray, concatenated along a new
axis (the new dimension 0).</p></td></tr></table></div><divclass="doc"><p>Gather specific elements from the TensorArray into output <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>.</p><p>All elements selected by <code>indices</code> must have the same shape.</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArrayReadV2"class="def">tensorArrayReadV2</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArrayReadV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a TensorArray.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>index</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong>: A float scalar that enforces proper chaining of operations.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype</td><tdclass="doc"><p><strong>value</strong>: The tensor that is read from the TensorArray.</p></td></tr></table></div><divclass="doc"><p>Read an element from the TensorArray into output <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>.</p></div></div><divclass="top"><pclass="src"><aname="v:decodeBase64"class="def">decodeBase64</a><ahref="src/TensorFlow-GenOps-Core.html#decodeBase64"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>input</strong>: Base64 strings to decode.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>output</strong>: Decoded strings.</p></td></tr></table></div><divclass="doc"><p>Decode web-safe base64-encoded strings.</p><p>Input may or may not have padding at the end. See EncodeBase64 for padding.
Web-safe means that input must use - and _ instead of + and /.</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArrayWriteV2"class="def">tensorArrayWriteV2</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArrayWriteV2"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a TensorArray.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>index</strong>: The position to write to inside the TensorArray.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>value</strong>: The tensor to write to the TensorArray.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong>: A float scalar that enforces proper chaining of operations.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_out</strong>: A float scalar that enforces proper chaining of operations.</p></td></tr></table></div><divclass="doc"><p>Push an element onto the tensor_array.</p></div></div><divclass="top"><pclass="src"><aname="v:audioSummary"class="def">audioSummary</a><ahref="src/TensorFlow-GenOps-Core.html#audioSummary"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>sample_rate</strong>: The sample rate of the signal in hertz.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>tag</strong>: Scalar. Used to build the <code>tag</code> attribute of the summary values.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>tensor</strong>: 2-D of shape `[batch_size, frames]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>summary</strong>: Scalar. Serialized <code>Summary</code> protocol buffer.</p></td></tr></table></div><divclass="doc"><p>Outputs a <code>Summary</code> protocol buffer with audio.</p><p>The summary has up to <code>max_outputs</code> summary values containing audio. The
audio is built from <code>tensor</code> which must be 3-D with shape `[batch_size,
frames, channels]` or 2-D with shape `[batch_size, frames]`. The values are
assumed to be in the range of `[-1.0, 1.0]` with a sample rate of <code>sample_rate</code>.</p><p>The <code>tag</code> argument is a scalar <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> of type <code>string</code>. It is used to
build the <code>tag</code> of the summary values:</p><ul><li>If <code>max_outputs</code> is 1, the summary value tag is '*tag*/audio'.</li><li>If <code>max_outputs</code> is greater than 1, the summary value tags are
generated sequentially as '*tag*/audio/0', '*tag*/audio/1', etc.</li></ul></div></div><divclass="top"><pclass="src"><aname="v:isFinite"class="def">isFinite</a><ahref="src/TensorFlow-GenOps-Core.html#isFinite"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Returns which elements of x are finite.</p><p><code>compatibility(numpy)
Equivalent to np.isfinite
</code>end_compatibility</p></div></div><divclass="top"><pclass="src"><aname="v:tensorArrayConcat"class="def">tensorArrayConcat</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArrayConcat"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p>(<strong>value</strong>, <strong>lengths</strong>)</p><ul><li><strong>value</strong></li><li><strong>lengths</strong></li></ul></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:sparseReduceSum"class="def">sparseReduceSum</a><ahref="src/TensorFlow-GenOps-Core.html#sparseReduceSum"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>input_indices</strong>: 2-D. `N x R` matrix with the indices of non-empty values in a
SparseTensor, possibly not in canonical ordering.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>input_values</strong>: 1-D. <code>N</code> non-empty values corresponding to <code>input_indices</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>input_shape</strong>: 1-D. Shape of the input SparseTensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>reduction_axes</strong>: 1-D. Length-<code>K</code> vector containing the reduction axes.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: `R-K`-D. The reduced Tensor.</p></td></tr></table></div><divclass="doc"><p>Computes the sum of elements across dimensions of a SparseTensor.</p><p>This Op takes a SparseTensor and is the sparse counterpart to
`tf.reduce_sum()`. In particular, this Op also returns a dense <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code>
instead of a sparse one.</p><p>Reduces <code>sp_input</code> along the dimensions given in <code>reduction_axes</code>. Unless
<code>keep_dims</code> is true, the rank of the tensor is reduced by 1 for each entry in
<code>reduction_axes</code>. If <code>keep_dims</code> is true, the reduced dimensions are retained
with length 1.</p><p>If <code>reduction_axes</code> has no entries, all dimensions are reduced, and a tensor
with a single element is returned. Additionally, the axes can be negative,
which are interpreted according to the indexing rules in Python.</p></div></div><divclass="top"><pclass="src"><aname="v:realDiv"class="def">realDiv</a><ahref="src/TensorFlow-GenOps-Core.html#realDiv"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns x / y element-wise for real types.</p><p>If <code>x</code> and <code>y</code> are reals, this will return the floating-point division.</p><ul><li>NOTE*: <code>Div</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:tensorArraySize"class="def">tensorArraySize</a><ahref="src/TensorFlow-GenOps-Core.html#tensorArraySize"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>flow_in</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>)</td><tdclass="doc"><p><strong>size</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:biasAddV1"class="def">biasAddV1</a><ahref="src/TensorFlow-GenOps-Core.html#biasAddV1"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>value</strong>: Any number of dimensions.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>bias</strong>: 1-D with size the last dimension of <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Broadcasted sum of <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code> and <code>bias</code>.</p></td></tr></table></div><divclass="doc"><p>Adds <code>bias</code> to <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code>.</p><p>This is a deprecated version of BiasAdd and will be soon removed.</p><p>This is a special case of `tf.add` where <code>bias</code> is restricted to be 1-D.
Broadcasting is supported, so <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#v:value">value</a></code> may have any number of dimensions.</p></div></div><divclass="top"><pclass="src"><aname="v:logicalOr"class="def">logicalOr</a><ahref="src/TensorFlow-GenOps-Core.html#logicalOr"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns the truth value of x OR y element-wise.</p><ul><li>NOTE*: <code>LogicalOr</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:stackPush"class="def">stackPush</a><ahref="src/TensorFlow-GenOps-Core.html#stackPush"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a stack.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>elem</strong>: The tensor to be pushed onto the stack.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p><strong>output</strong>: The same tensor as the input <code><ahref="../base-4.8.2.0/Data-Foldable.html#v:elem">elem</a></code>.</p></td></tr></table></div><divclass="doc"><p>Push an element onto the stack.</p></div></div><divclass="top"><pclass="src"><aname="v:quantizedRelu"class="def">quantizedRelu</a><ahref="src/TensorFlow-GenOps-Core.html#quantizedRelu"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tinput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` tinput, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` out_type)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 tinput</td><tdclass="doc"><p><strong>features</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>min_features</strong>: The float value that the lowest quantized value represents.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>max_features</strong>: The float value that the highest quantized value represents.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> out_type, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>activations</strong>, <strong>min_activations</strong>, <strong>max
sorted in the same order that they occur in <code>x</code>. This operation also returns a
tensor <code>idx</code> the same size as <code>x</code> that contains the index of each value of <code>x</code>
in the unique output <code>y</code>. Finally, it returns a third tensor <code>count</code> that
contains the count of each element of <code>y</code> in <code>x</code>. In other words:</p><p>`y[idx[i]] = x[i] for i in [0, 1,...,rank(x) - 1]`</p><p>For example:</p><p>```prettyprint
```</p></div></div><divclass="top"><pclass="src"><aname="v:truncateMod"class="def">truncateMod</a><ahref="src/TensorFlow-GenOps-Core.html#truncateMod"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Returns element-wise remainder of division. This emulates C semantics where</p><p>true, this follows C semantics in that the result here is consistent
with a flooring divide. E.g. `floor(x / y) * y + mod(x, y) = x`.</p><ul><li>NOTE*: <code>Mod</code> supports broadcasting. More about broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li></ul></div></div><divclass="top"><pclass="src"><aname="v:stridedSliceGrad"class="def">stridedSliceGrad</a><ahref="src/TensorFlow-GenOps-Core.html#stridedSliceGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> index, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` index)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 index</td><tdclass="doc"><p><strong>shape</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 index</td><tdclass="doc"><p><strong>begin</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 index</td><tdclass="doc"><p><strong>end</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 index</td><tdclass="doc"><p><strong>strides</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>dy</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Returns the gradient of <code>StridedSlice</code>.</p><p>Since <code>StridedSlice</code> cuts out pieces of its <code>input</code> which is size
<code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code>, its gradient will have the same shape (which is passed here
as <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code>). The gradient will be zero in any element that the slice
does not select.</p><p>Arguments are the same as StridedSliceGrad with the exception that
<code>dy</code> is the input gradient to be propagated and <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> is the
shape of <code>StridedSlice</code>'s <code>input</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:fractionalAvgPool"class="def">fractionalAvgPool</a><ahref="src/TensorFlow-GenOps-Core.html#fractionalAvgPool"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>value</strong>: 4-D with shape `[batch, height, width, channels]`.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p>(<strong>output</strong>, <strong>row_pooling_sequence</strong>, <strong>col_pooling_sequence</strong>)</p><ul><li><strong>output</strong>: output tensor after fractional avg pooling.</li><li><strong>row_pooling_sequence</strong>: row pooling sequence, needed to calculate gradient.</li><li><strong>col_pooling_sequence</strong>: column pooling sequence, needed to calculate gradient.</li></ul></td></tr></table></div><divclass="doc"><p>Performs fractional average pooling on the input.</p><p>Fractional average pooling is similar to Fractional max pooling in the pooling
region generation step. The only difference is that after pooling regions are
generated, a mean operation is performed instead of a max operation in each
pooling region.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseAccumulatorTakeGradient"class="def">sparseAccumulatorTakeGradient</a><ahref="src/TensorFlow-GenOps-Core.html#sparseAccumulatorTakeGradient"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` dtype)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to a SparseConditionalAccumulator.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>num_required</strong>: Number of gradients required before we return an aggregate.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>)</td><tdclass="doc"><p>(<strong>indices</strong>, <strong>values</strong>, <strong>shape</strong>)</p><ul><li><strong>indices</strong>: Indices of the average of the accumulated sparse gradients.</li><li><strong>values</strong>: Values of the average of the accumulated sparse gradients.</li><li><strong>shape</strong>: Shape of the average of the accumulated sparse gradients.</li></ul></td></tr></table></div><divclass="doc"><p>Extracts the average sparse gradient in the given SparseConditionalAccumulator,</p><p>provided that sufficient (i.e., more than num_required) gradients have been
accumulated. The op will blocks until sufficient gradients have been
accumulated. If the accumulator has already aggregated more than num_required
gradients, it will return its average of the accumulated gradients.
Also automatically increments the recorded global_step in the accumulator by 1,
and resets the aggregate to 0.</p></div></div><divclass="top"><pclass="src"><aname="v:decodeJSONExample"class="def">decodeJSONExample</a><ahref="src/TensorFlow-GenOps-Core.html#decodeJSONExample"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>json_examples</strong>: Each string is a JSON object serialized according to the JSON
mapping of the Example proto.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>binary_examples</strong>: Each string is a binary Example protocol buffer corresponding
to the respective element of <code>json_examples</code>.</p></td></tr></table></div><divclass="doc"><p>Convert JSON-encoded Example records to binary protocol buffer strings.</p><p>This op translates a tensor containing Example records, encoded using
the <ahref="https://developers.google.com/protocol-buffers/docs/proto3#json">standard JSON
mapping</a>,
into a tensor containing the same records encoded as binary protocol
buffers. The resulting tensor can then be fed to any of the other
Example-parsing ops.</p></div></div><divclass="top"><pclass="src"><aname="v:placeholderWithDefault"class="def">placeholderWithDefault</a><ahref="src/TensorFlow-GenOps-Core.html#placeholderWithDefault"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:Shape">Shape</a></td><tdclass="doc"><p><strong>shape</strong>: The (possibly partial) shape of the tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 dtype</td><tdclass="doc"><p><strong>input</strong>: The default value to produce when <code>output</code> is not fed.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype</td><tdclass="doc"><p><strong>output</strong>: A placeholder tensor that defaults to <code>input</code> if it is not fed.</p></td></tr></table></div><divclass="doc"><p>A placeholder op that passes though <code>input</code> when its output is not fed.</p></div></div><divclass="top"><pclass="src"><aname="v:applyFtrl"class="def">applyFtrl</a><ahref="src/TensorFlow-GenOps-Core.html#applyFtrl"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>var</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>accum</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>linear</strong>: Should be from a Variable().</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>grad</strong>: The gradient.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>lr</strong>: Scaling factor. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 t</td><tdclass="doc"><p><strong>l1</strong>: L1 regulariation. Must be a scalar.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v7 t</td><tdclass="doc"><p><strong>l2</strong>: L2 regulariation. Must be a scalar.</p></td></tr><tr><tdclass="src"
accum = accum_new</p></div></div><divclass="top"><pclass="src"><aname="v:sdcaShrinkL1"class="def">sdcaShrinkL1</a><ahref="src/TensorFlow-GenOps-Core.html#sdcaShrinkL1"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>l1</strong>: Symmetric l1 regularization strength.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>l2</strong>: Symmetric l2 regularization strength. Should be a positive float.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]</td><tdclass="doc"><p><strong>weights</strong>: a list of vectors where each value is the weight associated with a
feature group.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Applies L1 regularization shrink step on the parameters.</p></div></div><divclass="top"><pclass="src"><aname="v:shardedFilename"class="def">shardedFilename</a><ahref="src/TensorFlow-GenOps-Core.html#shardedFilename"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>basename</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>shard</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>num_shards</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>filename</strong></p></td></tr></table></div><divclass="doc"><p>Generate a sharded filename. The filename is printf formatted as</p><p>%s-%05d-of-%05d, basename, shard, num_shards.</p></div></div><divclass="top"><pclass="src"><aname="v:fakeQuantWithMinMaxArgs"class="def">fakeQuantWithMinMaxArgs</a><ahref="src/TensorFlow-GenOps-Core.html#fakeQuantWithMinMaxArgs"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>inputs</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>outputs</strong></p></td></tr></table></div><divclass="doc"><p>Fake-quantize the <code>inputs</code> tensor, type float to <code>outputs</code> tensor of same type.</p><p>Attributes [min; max] define the clamping range for the <code>inputs</code> data. Op
divides this range into 255 steps (total of 256 values), then replaces each
<code>inputs</code> value with the closest of the quantized step values.</p><p>Quantization is called fake since the output is still in floating point.</p></div></div><divclass="top"><pclass="src"><aname="v:scatterNdAdd"class="def">scatterNdAdd</a><ahref="src/TensorFlow-GenOps-Core.html#scatterNdAdd"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ref</strong>: A mutable Tensor. Should be from a Variable node.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>indices</strong>: A Tensor. Must be one of the following types: int32, int64.
A tensor of indices into ref.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>updates</strong>: A Tensor. Must have the same type as ref. A tensor of updated values
to add to ref.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output_ref</strong>: Same as ref. Returned as a convenience for operations that want
to use the updated values after the update is done.</p></td></tr></table></div><divclass="doc"><p>Applies sparse addition between <code>updates</code> and individual values or slices</p><p>within a given variable according to <code>indices</code>.</p><p><code>ref</code> is a <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> with rank <code>P</code> and <code>indices</code> is a <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> of rank <code>Q</code>.</p><p><code>indices</code> must be integer tensor, containing indices into <code>ref</code>.
It must be shape `[d_0, ..., d_{Q-2}, K]` where `0 < K <= P`.</p><p>The innermost dimension of <code>indices</code> (with length <code>K</code>) corresponds to
indices into elements (if `K = P`) or slices (if `K < P`) along the <code>K</code>th
dimension of <code>ref</code>.</p><p><code>updates</code> is <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> of rank `Q-1+P-K` with shape:</p><p>```
```</p><p>For example, say we want to add 4 scattered elements to a rank-1 tensor to 8
elements. In Python, that addition would look like this:</p><p>ref = tf.Variable([1, 2, 3, 4, 5, 6, 7, 8])
indices = tf.constant([[4], [3], [1], [7]])
updates = tf.constant([9, 10, 11, 12])
add = tf.scatter_nd_add(ref, indices, updates)
with tf.Session() as sess:
print sess.run(add)</p><p>The resulting update to ref would look like this:</p><dl><dt>1, 13, 3, 14, 14, 6, 7, 20</dt><dd></dd></dl><p>See <ahref="#scatter_nd">tf.scatter_nd</a> for more details about how to make updates to
slices.</p></div></div><divclass="top"><pclass="src"><aname="v:accumulatorNumAccumulated"class="def">accumulatorNumAccumulated</a><ahref="src/TensorFlow-GenOps-Core.html#accumulatorNumAccumulated"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>handle</strong>: The handle to an accumulator.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>)</td><tdclass="doc"><p><strong>num_accumulated</strong>: The number of gradients aggregated in the given accumulator.</p></td></tr></table></div><divclass="doc"><p>Returns the number of gradients aggregated in the given accumulators.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseSegmentSqrtN"class="def">sparseSegmentSqrtN</a><ahref="src/TensorFlow-GenOps-Core.html#sparseSegmentSqrtN"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>data</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>indices</strong>: A 1-D tensor. Has same rank as <code>segment_ids</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>segment_ids</strong>: A 1-D tensor. Values should be sorted and can be repeated.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Has same shape as data, except for dimension 0 which
has size <code>k</code>, the number of segments.</p></td></tr></table></div><divclass="doc"><p>Computes the sum along sparse segments of a tensor divided by the sqrt of N.</p><p>N is the size of the segment being reduced.</p><p>Read <ahref="../../api_docs/python/math_ops.md#segmentation">the section on
Segmentation</a> for an explanation
of segments.</p></div></div><divclass="top"><pclass="src"><aname="v:depthToSpace"class="def">depthToSpace</a><ahref="src/TensorFlow-GenOps-Core.html#depthToSpace"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>block_size</strong>: The size of the spatial block, same as in Space2Depth.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>DepthToSpace for tensors of type T.</p><p>Rearranges data from depth into blocks of spatial data.
This is the reverse transformation of SpaceToDepth. More specifically,
this op outputs a copy of the input tensor where values from the <code>depth</code>
dimension are moved in spatial blocks to the <code>height</code> and <code>width</code> dimensions.
The attr <code>block_size</code> indicates the input block size and how the data is moved.</p><ul><li>Chunks of data of size `block_size * block_size` from depth are rearranged
into non-overlapping blocks of size `block_size x block_size`</li><li>The width the output tensor is `input_depth * block_size`, whereas the
height is `input_height * block_size`.</li><li>The depth of the input tensor must be divisible by
`block_size * block_size`.</li></ul><p>That is, assuming the input is in the shape:
`[batch, height, width, depth]`,
the shape of the output will be:
`[batch, height*block_size, width*block_size, depth/(block_size*block_size)]`</p><p>This operation requires that the input tensor be of rank 4, and that
<code>block_size</code> be >=1 and that `block_size * block_size` be a divisor of the
input depth.</p><p>This operation is useful for resizing the activations between convolutions
(but keeping all data), e.g. instead of pooling. It is also useful for training
purely convolutional models.</p><p>For example, given this input of shape `[1, 1, 1, 4]`, and a block size of 2:</p><p>```prettyprint
x = [[[[1, 2, 3, 4]]]]</p><p>```</p><p>This operation will output a tensor of shape `[1, 2, 2, 1]`:</p><p>```prettyprint
[[[[1], [2]],
[[3], [4]]]]
```</p><p>Here, the input has a batch of 1 and each batch element has shape `[1, 1, 4]`,
the corresponding output will have 2x2 elements and will have a depth of
1 channel (1 = `4 / (block_size * block_size)`).
The output element shape is `[2, 2, 1]`.</p><p>For an input tensor with larger depth, here of shape `[1, 1, 1, 12]`, e.g.</p><p>```prettyprint
x = [[[[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12]]]]
```</p><p>This operation, for block size of 2, will return the following tensor of shape
`[1, 2, 2, 3]`</p><p>```prettyprint
[[[[1, 2, 3], [4, 5, 6]],
[[7, 8, 9], [10, 11, 12]]]]</p><p>```</p><p>Similarly, for the following input of shape `[1 2 2 4]`, and a block size of 2:</p><p>```prettyprint
x = [[[[1, 2, 3, 4],
[5, 6, 7, 8]],
[[9, 10, 11, 12],
[13, 14, 15, 16]]]]
```</p><p>the operator will return the following tensor of shape `[1 4 4 1]`:</p><p>```prettyprint
x = [[ [1], [2], [5], [6]],
[ [3], [4], [7], [8]],
[ [9], [10], [13], [14]],
[ [11], [12], [15], [16]]]</p><p>```</p></div></div><divclass="top"><pclass="src"><aname="v:allCandidateSampler"class="def">allCandidateSampler</a><ahref="src/TensorFlow-GenOps-Core.html#allCandidateSampler"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_sampled</strong>: Number of candidates to produce per batch.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_true</strong>: Number of true labels per context.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>unique</strong>: If unique is true, we sample with rejection, so that all sampled
candidates in a batch are unique. This requires some approximation to
estimate the post-rejection sampling probabilities.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>true_classes</strong>: A batch_size * num_true matrix, in which each row contains the
IDs of the num_true target_classes in the corresponding original label.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>sampled_candidates</strong>, <strong>true_expected_count</strong>, <strong>sampled_expected_count</strong>)</p><ul><li><strong>sampled_candidates</strong>: A vector of length num_sampled, in which each element is
the ID of a sampled candidate.</li><li><strong>true_expected_count</strong>: A batch_size * num_true matrix, representing
the number of times each candidate is expected to occur in a batch
of sampled candidates. If unique=true, then this is a probability.</li><li><strong>sampled_expected_count</strong>: A vector of length num_sampled, for each sampled
candidate representing the number of times the candidate is expected
to occur in a batch of sampled candidates. If unique=true, then this is a
probability.</li></ul></td></tr></table></div><divclass="doc"><p>Generates labels for candidate sampling with a learned unigram distribution.</p><p>See explanations of candidate sampling and the data formats at
go/candidate-sampling.</p><p>For each batch, this op picks a single set of sampled candidate labels.</p><p>The advantages of sampling candidates per-batch are simplicity and the
possibility of efficient dense matrix multiplication. The disadvantage is that
the sampled candidates must be chosen independently of the context and of the
true labels.</p></div></div><divclass="top"><pclass="src"><aname="v:resizeNearestNeighborGrad"class="def">resizeNearestNeighborGrad</a><ahref="src/TensorFlow-GenOps-Core.html#resizeNearestNeighborGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>grads</strong>: 4-D with shape `[batch, height, width, channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>size</strong>: = A 1-D int32 Tensor of 2 elements: `orig_height, orig_width`. The
original input size.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 4-D with shape `[batch, orig_height, orig_width, channels]`. Gradients
with respect to the input image.</p></td></tr></table></div><divclass="doc"><p>Computes the gradient of nearest neighbor interpolation.</p></div></div><divclass="top"><pclass="src"><aname="v:cTCGreedyDecoder"class="def">cTCGreedyDecoder</a><ahref="src/TensorFlow-GenOps-Core.html#cTCGreedyDecoder"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>inputs</strong>: 3-D, shape: `(max_time x batch_size x num_classes)`, the logits.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>sequence_length</strong>: A vector containing sequence lengths, size `(batch_size)`.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>decoded_indices</strong>, <strong>decoded_values</strong>, <strong>decoded_shape</strong>, <strong>log_probability</strong>)</p><ul><li><strong>decoded_indices</strong>: Indices matrix, size `(total_decoded_outputs x 2)`,
of a `SparseTensor<ahref="int64,">2</a>`. The rows store: [batch, time].</li><li><strong>decoded_values</strong>: Values vector, size: `(total_decoded_outputs)`,
of a `SparseTensor<ahref="int64,">2</a>`. The vector stores the decoded classes.</li><li><strong>decoded_shape</strong>: Shape vector, size `(2)`, of the decoded SparseTensor.
log-probabilities.</li></ul></td></tr></table></div><divclass="doc"><p>Performs greedy decoding on the logits given in inputs.</p><p>A note about the attribute merge_repeated: if enabled, when
consecutive logits' maximum indices are the same, only the first of
these is emitted. Labeling the blank <code><ahref="../base-4.8.2.0/Prelude.html#v:-42-">*</a></code>, the sequence "A B B * B B"
becomes "A B" if merge_repeated = True and "A B B B B" if
merge_repeated = False.</p><p>Regardless of the value of merge_repeated, if the maximum index of a given
time and batch corresponds to the blank, index `(num_classes - 1)`, no new
element is emitted.</p></div></div><divclass="top"><pclass="src"><aname="v:l2Loss"class="def">l2Loss</a><ahref="src/TensorFlow-GenOps-Core.html#l2Loss"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>t</strong>: Typically 2-D, but may have any dimensions.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 0-D.</p></td></tr></table></div><divclass="doc"><p>L2 Loss.</p><p>Computes half the L2 norm of a tensor without the <code><ahref="../base-4.8.2.0/Prelude.html#v:sqrt">sqrt</a></code>:</p><p>output = sum(t ** 2) / 2</p></div></div><divclass="top"><pclass="src"><aname="v:segmentMax"class="def">segmentMax</a><ahref="src/TensorFlow-GenOps-Core.html#segmentMax"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>data</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>segment_ids</strong>: A 1-D tensor whose rank is equal to the rank of `data`'s
first dimension. Values should be sorted and can be repeated.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Has same shape as data, except for dimension 0 which
has size <code>k</code>, the number of segments.</p></td></tr></table></div><divclass="doc"><p>Computes the maximum along segments of a tensor.</p><p>Read <ahref="../../api_docs/python/math_ops.md#segmentation">the section on Segmentation</a>
for an explanation of segments.</p><p>Computes a tensor such that
\(output_i = max_j(data_j)\) where <code><ahref="../base-4.8.2.0/Data-Ord.html#v:max">max</a></code> is over <code>j</code> such
<ahref="/div">/div</a></p></div></div><divclass="top"><pclass="src"><aname="v:countUpTo"class="def">countUpTo</a><ahref="src/TensorFlow-GenOps-Core.html#countUpTo"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>limit</strong>: If incrementing ref would bring it above limit, instead generates an
<code>OutOfRange</code> error.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ref</strong>: Should be from a scalar <code>Variable</code> node.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p><strong>output</strong>: A copy of the input before increment. If nothing else modifies the
input, the values produced will all be distinct.</p></td></tr></table></div><divclass="doc"><p>Increments <code>ref</code> until it reaches <code>limit</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:tFRecordReader"class="def">tFRecordReader</a><ahref="src/TensorFlow-GenOps-Core.html#tFRecordReader"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>)</td><tdclass="doc"><p><strong>reader_handle</strong>: The handle to reference the Reader.</p></td></tr></table></div><divclass="doc"><p>A Reader that outputs the records from a TensorFlow Records file.</p></div></div><divclass="top"><pclass="src"><aname="v:switch"class="def">switch</a><ahref="src/TensorFlow-GenOps-Core.html#switch"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>data</strong>: The tensor to be forwarded to the appropriate output.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>pred</strong>: A scalar that specifies which output port will receive data.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p>(<strong>output_false</strong>, <strong>output_true</strong>)</p><ul><li><strong>output_false</strong>: If <code><ahref="../base-4.8.2.0/Prelude.html#v:pred">pred</a></code> is false, data will be forwarded to this output.</li><li><strong>output_true</strong>: If <code><ahref="../base-4.8.2.0/Prelude.html#v:pred">pred</a></code> is true, data will be forwarded to this output.</li></ul></td></tr></table></div><divclass="doc"><p>Forwards `data` to the output port determined by <code><ahref="../base-4.8.2.0/Prelude.html#v:pred">pred</a></code>.</p><p>If <code><ahref="../base-4.8.2.0/Prelude.html#v:pred">pred</a></code> is true, the `data` input is forwarded to <code>output_true</code>. Otherwise,
the data goes to <code>output_false</code>.</p><p>See also <code>RefSwitch</code> and <code>Merge</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseSegmentMeanGrad"class="def">sparseSegmentMeanGrad</a><ahref="src/TensorFlow-GenOps-Core.html#sparseSegmentMeanGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>grad</strong>: gradient propagated to the SparseSegmentMean op.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>indices</strong>: indices passed to the corresponding SparseSegmentMean op.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>segment_ids</strong>: segment_ids passed to the corresponding SparseSegmentMean op.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>output_dim0</strong>: dimension 0 of "data" passed to SparseSegmentMean op.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Computes gradients for SparseSegmentMean.</p><p>Returns tensor "output" with same shape as grad, except for dimension 0 whose
value is output_dim0.</p></div></div><divclass="top"><pclass="src"><aname="v:gatherNd"class="def">gatherNd</a><ahref="src/TensorFlow-GenOps-Core.html#gatherNd"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tparams, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 tparams</td><tdclass="doc"><p><strong>params</strong>: `P-D`. The tensor from which to gather values.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>indices</strong>: `Q-D`. Index tensor having shape `[d_0, ..., d_{Q-2}, K]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> tparams</td><tdclass="doc"><p><strong>output</strong>: `(P+Q-K-1)-D`. Values from <code>params</code> gathered from indices given by
<code>indices</code>.</p></td></tr></table></div><divclass="doc"><p>Gather values or slices from <code>params</code> according to <code>indices</code>.</p><p><code>params</code> is a Tensor of rank <code>P</code> and <code>indices</code> is a Tensor of rank <code>Q</code>.</p><p><code>indices</code> must be integer tensor, containing indices into <code>params</code>.
It must be shape `[d_0, ..., d_{Q-2}, K]` where `0 < K <= P`.</p><p>The innermost dimension of <code>indices</code> (with length <code>K</code>) corresponds to
indices into elements (if `K = P`) or slices (if `K < P`) along the <code>K</code>th
dimension of <code>params</code>.</p><p>Produces an output tensor with shape</p><p>```
```</p></div></div><divclass="top"><pclass="src"><aname="v:squeeze"class="def">squeeze</a><ahref="src/TensorFlow-GenOps-Core.html#squeeze"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: The <code>input</code> to squeeze.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: Contains the same data as <code>input</code>, but has one or more dimensions of
size 1 removed.</p></td></tr></table></div><divclass="doc"><p>Removes dimensions of size 1 from the shape of a tensor.</p><p>Given a tensor <code>input</code>, this operation returns a tensor of the same type with
all dimensions of size 1 removed. If you don't want to remove all size 1
dimensions, you can remove specific size 1 dimensions by specifying
# <code>t</code> is a tensor of shape [1, 2, 1, 3, 1, 1]
shape(squeeze(t)) ==> [2, 3]
```</p><p>Or, to remove specific size 1 dimensions:</p><p>```prettyprint
# <code>t</code> is a tensor of shape [1, 2, 1, 3, 1, 1]
shape(squeeze(t, [2, 4])) ==> [1, 2, 3, 1]
```</p></div></div><divclass="top"><pclass="src"><aname="v:randomUniform"class="def">randomUniform</a><ahref="src/TensorFlow-GenOps-Core.html#randomUniform"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` dtype, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>shape</strong>: The shape of the output tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> dtype)</td><tdclass="doc"><p><strong>output</strong>: A tensor of the specified shape filled with uniform random values.</p></td></tr></table></div><divclass="doc"><p>Outputs random values from a uniform distribution.</p><p>The generated values follow a uniform distribution in the range `[0, 1)`. The
lower bound 0 is included in the range, while the upper bound 1 is excluded.</p></div></div><divclass="top"><pclass="src"><aname="v:readerReadUpTo"class="def">readerReadUpTo</a><ahref="src/TensorFlow-GenOps-Core.html#readerReadUpTo"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>reader_handle</strong>: Handle to a <code>Reader</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>queue_handle</strong>: Handle to a <code>Queue</code>, with string work items.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_records</strong>: number of records to read from <code>Reader</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>)</td><tdclass="doc"><p>(<strong>keys</strong>, <strong>values</strong>)</p><ul><li><strong>keys</strong>: A 1-D tensor.</li><li><strong>values</strong>: A 1-D tensor.</li></ul></td></tr></table></div><divclass="doc"><p>Returns up to <code>num_records</code> (key, value) pairs produced by a Reader.</p><p>Will dequeue from the input queue if necessary (e.g. when the
Reader needs to start reading from a new file since it has finished
with the previous file).
It may return less than <code>num_records</code> even before the last batch.</p></div></div><divclass="top"><pclass="src"><aname="v:conv3DBackpropInput"class="def">conv3DBackpropInput</a><ahref="src/TensorFlow-GenOps-Core.html#conv3DBackpropInput"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Shape `[batch, depth, rows, cols, in_channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>filter</strong>: Shape `[depth, rows, cols, in_channels, out_channels]`.
<code>in_channels</code> must match between <code>input</code> and <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>out_backprop</strong>: Backprop signal of shape `[batch, out_depth, out_rows, out_cols,
out_channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Computes the gradients of 3-D convolution with respect to the input.</p></div></div><divclass="top"><pclass="src"><aname="v:depthwiseConv2dNative"class="def">depthwiseConv2dNative</a><ahref="src/TensorFlow-GenOps-Core.html#depthwiseConv2dNative"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>filter</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Computes a 2-D depthwise convolution given 4-D <code>input</code> and <code><ahref="../base-4.8.2.0/GHC-OldList.html#v:filter">filter</a></code> tensors.</p><p>Given an input tensor of shape `[batch, in_height, in_width, in_channels]`
filter[di, dj, k, q]</p><p>Must have `strides[0] = strides[3] = 1`. For the most common case of the same
horizontal and vertices strides, `strides = [1, stride, stride, 1]`.</p></div></div><divclass="top"><pclass="src"><aname="v:learnedUnigramCandidateSampler"class="def">learnedUnigramCandidateSampler</a><ahref="src/TensorFlow-GenOps-Core.html#learnedUnigramCandidateSampler"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_sampled</strong>: Number of candidates to randomly sample per batch.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_true</strong>: Number of true labels per context.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>range_max</strong>: The sampler will sample integers from the interval [0, range_max).</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>unique</strong>: If unique is true, we sample with rejection, so that all sampled
candidates in a batch are unique. This requires some approximation to
estimate the post-rejection sampling probabilities.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>true_classes</strong>: A batch_size * num_true matrix, in which each row contains the
IDs of the num_true target_classes in the corresponding original label.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>sampled_candidates</strong>, <strong>true_expected_count</strong>, <strong>sampled_expected_count</strong>)</p><ul><li><strong>sampled_candidates</strong>: A vector of length num_sampled, in which each element is
the ID of a sampled candidate.</li><li><strong>true_expected_count</strong>: A batch_size * num_true matrix, representing
the number of times each candidate is expected to occur in a batch
of sampled candidates. If unique=true, then this is a probability.</li><li><strong>sampled_expected_count</strong>: A vector of length num_sampled, for each sampled
candidate representing the number of times the candidate is expected
to occur in a batch of sampled candidates. If unique=true, then this is a
probability.</li></ul></td></tr></table></div><divclass="doc"><p>Generates labels for candidate sampling with a learned unigram distribution.</p><p>See explanations of candidate sampling and the data formats at
go/candidate-sampling.</p><p>For each batch, this op picks a single set of sampled candidate labels.</p><p>The advantages of sampling candidates per-batch are simplicity and the
possibility of efficient dense matrix multiplication. The disadvantage is that
the sampled candidates must be chosen independently of the context and of the
true labels.</p></div></div><divclass="top"><pclass="src"><aname="v:initializeTable"class="def">initializeTable</a><ahref="src/TensorFlow-GenOps-Core.html#initializeTable"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tkey, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tval)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>table_handle</strong>: Handle to a table which will be initialized.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tkey</td><tdclass="doc"><p><strong>keys</strong>: Keys of type Tkey.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 tval</td><tdclass="doc"><p><strong>values</strong>: Values of type Tval.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Table initializer that takes two tensors for keys and values respectively.</p></div></div><divclass="top"><pclass="src"><aname="v:merge"class="def">merge</a><ahref="src/TensorFlow-GenOps-Core.html#merge"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t]</td><tdclass="doc"><p><strong>inputs</strong>: The input tensors, exactly one of which will become available.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>)</td><tdclass="doc"><p>(<strong>output</strong>, <strong>value_index</strong>)</p><ul><li><strong>output</strong>: Will be set to the available input tensor.</li><li><strong>value_index</strong>: The index of the chosen input tensor in <code>inputs</code>.</li></ul></td></tr></table></div><divclass="doc"><p>Forwards the value of an available tensor from <code>inputs</code> to <code>output</code>.</p><p><code>Merge</code> waits for at least one of the tensors in <code>inputs</code> to become available.
It is usually combined with <code>Switch</code> to implement branching.</p><p><code>Merge</code> forwards the first tensor for become available to <code>output</code>, and sets
<code>value_index</code> to its index in <code>inputs</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:refMerge"class="def">refMerge</a><ahref="src/TensorFlow-GenOps-Core.html#refMerge"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t]</td><tdclass="doc"><p><strong>inputs</strong>: The input tensors, exactly one of which will become available.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>)</td><tdclass="doc"><p>(<strong>output</strong>, <strong>value_index</strong>)</p><ul><li><strong>output</strong>: Will be set to the available input tensor.</li><li><strong>value_index</strong>: The index of the chosen input tensor in <code>inputs</code>.</li></ul></td></tr></table></div><divclass="doc"><p>Forwards the value of an available tensor from <code>inputs</code> to <code>output</code>.</p><p><code>Merge</code> waits for at least one of the tensors in <code>inputs</code> to become available.
It is usually combined with <code>Switch</code> to implement branching.</p><p><code>Merge</code> forwards the first tensor for become available to <code>output</code>, and sets
<code>value_index</code> to its index in <code>inputs</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:round"class="def">round</a><ahref="src/TensorFlow-GenOps-Core.html#round"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>y</strong></p></td></tr></table></div><divclass="doc"><p>Rounds the values of a tensor to the nearest integer, element-wise.</p><p>Rounds half to even. Also known as bankers rounding. If you want to round
according to the current system rounding mode use std::cint.</p></div></div><divclass="top"><pclass="src"><aname="v:batchSelfAdjointEig"class="def">batchSelfAdjointEig</a><ahref="src/TensorFlow-GenOps-Core.html#batchSelfAdjointEig"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div></div><divclass="top"><pclass="src"><aname="v:dynamicPartition"class="def">dynamicPartition</a><ahref="src/TensorFlow-GenOps-Core.html#dynamicPartition"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_partitions</strong>: The number of partitions to output.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>data</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>partitions</strong>: Any shape. Indices in the range `[0, num_partitions)`.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t]</td><tdclass="doc"><p><strong>outputs</strong></p></td></tr></table></div><divclass="doc"><p>Partitions `data` into <code>num_partitions</code> tensors using indices from <code>partitions</code>.</p><p>For each index tuple <code>js</code> of size `partitions.ndim`, the slice `data[js, ...]`
<ahref="/div">/div</a></p></div></div><divclass="top"><pclass="src"><aname="v:reshape"class="def">reshape</a><ahref="src/TensorFlow-GenOps-Core.html#reshape"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tshape, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tshape)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>tensor</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tshape</td><tdclass="doc"><p><strong>shape</strong>: Defines the shape of the output tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong></p></td></tr></table></div><divclass="doc"><p>Reshapes a tensor.</p><p>Given <code>tensor</code>, this operation returns a tensor that has the same values
as <code>tensor</code> with shape <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code>.</p><p>If one component of <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> is the special value -1, the size of that dimension
is computed so that the total size remains constant. In particular, a <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code>
of `[-1]` flattens into 1-D. At most one component of <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> can be -1.</p><p>If <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> is 1-D or higher, then the operation returns a tensor with shape
<code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> filled with the values of <code>tensor</code>. In this case, the number of elements
implied by <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> must be the same as the number of elements in <code>tensor</code>.</p><p>For example:</p><p>```prettyprint
[7, 8, 9]]</p><p># tensor <code>t</code> is [[[1, 1], [2, 2]],
# [[3, 3], [4, 4]]]
# tensor <code>t</code> has shape [2, 2, 2]
reshape(t, [2, 4]) ==> [[1, 1, 2, 2],
[3, 3, 4, 4]]</p><p># tensor <code>t</code> is [[[1, 1, 1],
# [2, 2, 2]],
# [[3, 3, 3],
# [4, 4, 4]],
# [[5, 5, 5],
# [6, 6, 6]]]
# tensor <code>t</code> has shape [3, 2, 3]
# pass '[-1]' to flatten <code>t</code>
reshape(t, [-1]) ==> [1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4, 4, 5, 5, 5, 6, 6, 6]</p><p># -1 can also be used to infer the shape</p><p># -1 is inferred to be 9:
```</p></div></div><divclass="top"><pclass="src"><aname="v:fixedLengthRecordReader"class="def">fixedLengthRecordReader</a><ahref="src/TensorFlow-GenOps-Core.html#fixedLengthRecordReader"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>record_bytes</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>)</td><tdclass="doc"><p><strong>reader_handle</strong>: The handle to reference the Reader.</p></td></tr></table></div><divclass="doc"><p>A Reader that outputs fixed-length records from a file.</p></div></div><divclass="top"><pclass="src"><aname="v:sdcaOptimizer"class="def">sdcaOptimizer</a><ahref="src/TensorFlow-GenOps-Core.html#sdcaOptimizer"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>l1</strong>: Symmetric l1 regularization strength.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>l2</strong>: Symmetric l2 regularization strength.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_inner_iterations</strong>: Number of iterations per mini-batch.</p></td></tr><tr><tdclass="src">-><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>num_loss_partitions</strong>: Number of partitions of the global loss function.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]</td><tdclass="doc"><p><strong>sparse_example_indices</strong>: a list of vectors which contain example indices.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]</td><tdclass="doc"><p><strong>sparse_feature_indices</strong>: a list of vectors which contain feature indices.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]</td><tdclass="doc"><p><strong>sparse_feature_values</strong>: a list of vectors which contains feature value
associated with each feature group.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]</td><tdclass="doc"><p><strong>dense_features</strong>: a list of matrices which contains the dense feature values.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>example_weights</strong>: a vector which contains the weight associated with each
example.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>example_labels</strong>: a vector which contains the label/target associated with each
example.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v7 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]</td><tdclass="doc"><p><strong>sparse_indices</strong>: a list of vectors where each value is the indices which has
corresponding weights in sparse_weights. This field maybe ommitted for the
dense approach.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v8 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]</td><tdclass="doc"><p><strong>sparse_weights</strong>: a list of vectors where each value is the weight associated with
a sparse feature group.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v9 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]</td><tdclass="doc"><p><strong>dense_weights</strong>: a list of vectors where the values are the weights associated
with a dense feature group.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v10 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>example_state_data</strong>: a list of vectors containing the example state data.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>], [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>])</td><tdclass="doc"><p>(<strong>out_example_state_data</strong>, <strong>out_delta_sparse_weights</strong>, <strong>out_delta_dense_weights</strong>)</p><ul><li><strong>out_example_state_data</strong>: a list of vectors containing the updated example state
data.</li><li><strong>out_delta_sparse_weights</strong>: a list of vectors where each value is the delta
weights associated with a sparse feature group.</li><li><strong>out_delta_dense_weights</strong>: a list of vectors where the values are the delta
weights associated with a dense feature group.</li></ul></td></tr></table></div><divclass="doc"><p>Distributed version of Stochastic Dual Coordinate Ascent (SDCA) optimizer for</p><p>linear models with L1 + L2 regularization. As global optimization objective is
strongly-convex, the optimizer optimizes the dual objective at each step. The
optimizer applies each update one example at a time. Examples are sampled
uniformly, and the optimizer is learning rate free and enjoys linear convergence
2012 arXiv1211.2717S: <ahref="http://arxiv.org/pdf/1211.2717v1.pdf">http://arxiv.org/pdf/1211.2717v1.pdf</a></p><p>Loss objective = sum f_{i}(wx_{i}) + (l2 / 2) * |w|^2 + l1 * |w|</p><p>Adding vs. Averaging in Distributed Primal-Dual Optimization.
Chenxin Ma, Virginia Smith, Martin Jaggi, Michael I. Jordan, Peter Richtarik,
Martin Takac <ahref="http://arxiv.org/abs/1502.03508">http://arxiv.org/abs/1502.03508</a></p><p>Stochastic Dual Coordinate Ascent with Adaptive Probabilities
Dominik Csiba, Zheng Qu, Peter Richtarik <ahref="https://arxiv.org/abs/1502.08053">https://arxiv.org/abs/1502.08053</a></p></div></div><divclass="top"><pclass="src"><aname="v:resizeArea"class="def">resizeArea</a><ahref="src/TensorFlow-GenOps-Core.html#resizeArea"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>images</strong>: 4-D with shape `[batch, height, width, channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>size</strong>: = A 1-D int32 Tensor of 2 elements: `new_height, new_width`. The
new size for the images.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>resized_images</strong>: 4-D with shape
`[batch, new_height, new_width, channels]`.</p></td></tr></table></div><divclass="doc"><p>Resize <code>images</code> to <code><ahref="TensorFlow-GenOps-Core.html#v:size">size</a></code> using area interpolation.</p><p>Input images can be of different types but output images are always float.</p></div></div><divclass="top"><pclass="src"><aname="v:linSpace"class="def">linSpace</a><ahref="src/TensorFlow-GenOps-Core.html#linSpace"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>start</strong>: First entry in the range.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>stop</strong>: Last entry in the range.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 tidx</td><tdclass="doc"><p><strong>num</strong>: Number of values to generate.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 1-D. The generated values.</p></td></tr></table></div><divclass="doc"><p>Generates values in an interval.</p><p>A sequence of <code>num</code> evenly-spaced values are generated beginning at <code>start</code>.
If `num > 1`, the values in the sequence increase by `stop - start / num - 1`,
so that the last one is exactly <code>stop</code>.</p><p>For example:</p><p>```
```</p></div></div><divclass="top"><pclass="src"><aname="v:cTCLoss"class="def">cTCLoss</a><ahref="src/TensorFlow-GenOps-Core.html#cTCLoss"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a></td><tdclass="doc"><p><strong>inputs</strong>: 3-D, shape: `(max_time x batch_size x num_classes)`, the logits.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>labels_indices</strong>: The indices of a `SparseTensor<ahref="int32,">2</a>`.
`labels_indices(i, :) == [b, t]` means `labels_values(i)` stores the id for
`(batch b, time t)`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>labels_values</strong>: The values (labels) associated with the given batch and time.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>sequence_length</strong>: A vector containing sequence lengths (batch).</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>)</td><tdclass="doc"><p>(<strong>loss</strong>, <strong>gradient</strong>)</p><ul><li><strong>loss</strong>: A vector (batch) containing log-probabilities.</li><li><strong>gradient</strong>: The gradient of <code>loss</code>. 3-D, shape:
`(max_time x batch_size x num_classes)`.</li></ul></td></tr></table></div><divclass="doc"><p>Calculates the CTC Loss (log probability) for each batch entry. Also calculates</p><p>the gradient. This class performs the softmax operation for you, so inputs
should be e.g. linear projections of outputs by an LSTM.</p></div></div><divclass="top"><pclass="src"><aname="v:matrixDiagPart"class="def">matrixDiagPart</a><ahref="src/TensorFlow-GenOps-Core.html#matrixDiagPart"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: Rank <code>k</code> tensor where `k >= 2`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>diagonal</strong>: The extracted diagonal(s) having shape
`diagonal.shape = input.shape[:-2] + [min(input.shape[-2:])]`.</p></td></tr></table></div><divclass="doc"><p>Returns the batched diagonal part of a batched tensor.</p><p>This operation returns a tensor with the <code>diagonal</code> part
of the batched <code>input</code>. The <code>diagonal</code> part is computed as follows:</p><p>Assume <code>input</code> has <code>k</code> dimensions `[I, J, K, ..., M, N]`, then the output is a
tensor of rank `k - 1` with dimensions `[I, J, K, ..., min(M, N)]` where:</p><p>`diagonal[i, j, k, ..., n] = input[i, j, k, ..., n, n]`.</p><p>The input must be at least a matrix.</p><p>For example:</p><p>```prettyprint
```</p></div></div><divclass="top"><pclass="src"><aname="v:enter"class="def">enter</a><ahref="src/TensorFlow-GenOps-Core.html#enter"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>data</strong>: The tensor to be made available to the child frame.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The same tensor as `data`.</p></td></tr></table></div><divclass="doc"><p>Creates or finds a child frame, and makes `data` available to the child frame.</p><p>This op is used together with <code>Exit</code> to create loops in the graph.
The unique <code>frame_name</code> is used by the <code>Executor</code> to identify frames. If
<code>is_constant</code> is true, <code>output</code> is a constant in the child frame; otherwise
it may be changed in the child frame. At most <code>parallel_iterations</code> iterations
are run in parallel in the child frame.</p></div></div><divclass="top"><pclass="src"><aname="v:encodePng"class="def">encodePng</a><ahref="src/TensorFlow-GenOps-Core.html#encodePng"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>image</strong>: 3-D with shape `[height, width, channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>contents</strong>: 0-D. PNG-encoded image.</p></td></tr></table></div><divclass="doc"><p>PNG-encode an image.</p><p><code>image</code> is a 3-D uint8 or uint16 Tensor of shape `[height, width, channels]`
where <code>channels</code> is:</p><ul><li>1: for grayscale.</li><li>2: for grayscale + alpha.</li><li>3: for RGB.</li><li>4: for RGBA.</li></ul><p>The ZLIB compression level, <code>compression</code>, can be -1 for the PNG-encoder
default or a value from 0 to 9. 9 is the highest compression level, generating
the smallest output, but is slower.</p></div></div><divclass="top"><pclass="src"><aname="v:exit"class="def">exit</a><ahref="src/TensorFlow-GenOps-Core.html#exit"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>data</strong>: The tensor to be made available to the parent frame.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The same tensor as `data`.</p></td></tr></table></div><divclass="doc"><p>Exits the current frame to its parent frame.</p><p>Exit makes its input `data` available to the parent frame.</p></div></div><divclass="top"><pclass="src"><aname="v:scatterNd"class="def">scatterNd</a><ahref="src/TensorFlow-GenOps-Core.html#scatterNd"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 tindices</td><tdclass="doc"><p><strong>indices</strong>: A Tensor. Must be one of the following types: int32, int64.
A tensor of indices into ref.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>updates</strong>: A Tensor. Must have the same type as tensor. A tensor of updated values
to store in ref.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 tindices</td><tdclass="doc"><p><strong>shape</strong>: A vector. The shape of the resulting tensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: A new tensor with the given shape and updates applied according
to the indices.</p></td></tr></table></div><divclass="doc"><p>Creates a new tensor by applying sparse <code>updates</code> to individual</p><p>values or slices within a zero tensor of the given <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> tensor according to
indices. This operator is the inverse of the <ahref="#gather_nd">tf.gather_nd</a>
operator which extracts values or slices from a given tensor.</p><p>TODO(simister): Add a link to Variable.<strong>getitem</strong> documentation on slice
syntax.</p><p><code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code> is a <code>TensorShape</code> with rank <code>P</code> and <code>indices</code> is a <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> of rank
<code>Q</code>.</p><p><code>indices</code> must be integer tensor, containing indices into <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code>.
It must be shape `[d_0, ..., d_{Q-2}, K]` where `0 < K <= P`.</p><p>The innermost dimension of <code>indices</code> (with length <code>K</code>) corresponds to
indices into elements (if `K = P`) or slices (if `K < P`) along the <code>K</code>th
dimension of <code><ahref="TensorFlow-GenOps-Core.html#v:shape">shape</a></code>.</p><p><code>updates</code> is Tensor of rank `Q-1+P-K` with shape:</p><p>```
[d_0, ..., d_{Q-2}, shape[K], ..., shape[P-1]].
```</p><p>The simplest form of scatter is to insert individual elements in a tensor by
index. For example, say we want to insert 4 scattered elements in a rank-1
tensor with 8 elements.</p><p><ahref="div">style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;"</a>
<ahref="/div">/div</a></p><p>In Python, this scatter operation would look like this:</p><p>indices = tf.constant([[4], [3], [1], [7]])
updates = tf.constant([9, 10, 11, 12])
shape = tf.constant([8])
scatter = tf.scatter_nd(indices, updates, shape)
with tf.Session() as sess:
print sess.run(scatter)</p><p>The resulting tensor would look like this:</p><dl><dt>0, 11, 0, 10, 9, 0, 0, 12</dt><dd></dd></dl><p>We can also, insert entire slices of a higher rank tensor all at once. For
example, if we wanted to insert two slices in the first dimension of a
rank-3 tensor with two matrices of new values.</p><p><ahref="div">style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;"</a>
print sess.run(scatter)</p><p>The resulting tensor would look like this:</p><dl><dt>[[5, 5, 5, 5</dt><dd>, [6, 6, 6, 6], [7, 7, 7, 7], [8, 8, 8, 8]],</dd><dt>[0, 0, 0, 0</dt><dd>, [0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0]],</dd><dt>[5, 5, 5, 5</dt><dd>, [6, 6, 6, 6], [7, 7, 7, 7], [8, 8, 8, 8]],</dd><dt>[0, 0, 0, 0</dt><dd>, [0, 0, 0, 0], [0, 0, 0, 0], [0, 0, 0, 0]]]</dd></dl></div></div><divclass="top"><pclass="src"><aname="v:priorityQueue"class="def">priorityQueue</a><ahref="src/TensorFlow-GenOps-Core.html#priorityQueue"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>)</td><tdclass="doc"><p><strong>handle</strong>: The handle to the queue.</p></td></tr></table></div><divclass="doc"><p>A queue that produces elements sorted by the first component value.</p><p>Note that the PriorityQueue requires the first component of any element
to be a scalar int64, in addition to the other elements declared by
component_types. Therefore calls to Enqueue and EnqueueMany (resp. Dequeue
and DequeueMany) on a PriorityQueue will all require (resp. output) one extra
entry in their input (resp. output) lists.</p></div></div><divclass="top"><pclass="src"><aname="v:refSwitch"class="def">refSwitch</a><ahref="src/TensorFlow-GenOps-Core.html#refSwitch"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>data</strong>: The ref tensor to be forwarded to the appropriate output.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Bool.html#t:Bool">Bool</a></td><tdclass="doc"><p><strong>pred</strong>: A scalar that specifies which output port will receive data.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p>(<strong>output_false</strong>, <strong>output_true</strong>)</p><ul><li><strong>output_false</strong>: If <code><ahref="../base-4.8.2.0/Prelude.html#v:pred">pred</a></code> is false, data will be forwarded to this output.</li><li><strong>output_true</strong>: If <code><ahref="../base-4.8.2.0/Prelude.html#v:pred">pred</a></code> is true, data will be forwarded to this output.</li></ul></td></tr></table></div><divclass="doc"><p>Forwards the ref tensor `data` to the output port determined by <code><ahref="../base-4.8.2.0/Prelude.html#v:pred">pred</a></code>.</p><p>If <code><ahref="../base-4.8.2.0/Prelude.html#v:pred">pred</a></code> is true, the `data` input is forwarded to <code>output_true</code>. Otherwise,
the data goes to <code>output_false</code>.</p><p>See also <code>Switch</code> and <code>Merge</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:nextIteration"class="def">nextIteration</a><ahref="src/TensorFlow-GenOps-Core.html#nextIteration"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>data</strong>: The tensor to be made available to the next iteration.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The same tensor as `data`.</p></td></tr></table></div><divclass="doc"><p>Makes its input available to the next iteration.</p></div></div><divclass="top"><pclass="src"><aname="v:refNextIteration"class="def">refNextIteration</a><ahref="src/TensorFlow-GenOps-Core.html#refNextIteration"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>data</strong>: The tensor to be made available to the next iteration.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output</strong>: The same tensor as `data`.</p></td></tr></table></div><divclass="doc"><p>Makes its input available to the next iteration.</p></div></div><divclass="top"><pclass="src"><aname="v:batchMatMul"class="def">batchMatMul</a><ahref="src/TensorFlow-GenOps-Core.html#batchMatMul"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong>: 3-D or higher with shape `[..., r_x, c_x]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong>: 3-D or higher with shape `[..., r_y, c_y]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 3-D or higher with shape `[..., r_o, c_o]`</p></td></tr></table></div><divclass="doc"><p>Multiplies slices of two tensors in batches.</p><p>Multiplies all slices of <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code><code>x</code> and <code>y</code> (each slice can be
viewed as an element of a batch), and arranges the individual results
in a single output tensor of the same batch size. Each of the
individual slices can optionally be adjointed (to adjoint a matrix
means to transpose and conjugate it) before multiplication by setting
the <code>adj_x</code> or <code>adj_y</code> flag to <code><ahref="../base-4.8.2.0/Data-Bool.html#v:True">True</a></code>, which are by default <code><ahref="../base-4.8.2.0/Data-Bool.html#v:False">False</a></code>.</p><p>The input tensors <code>x</code> and <code>y</code> are 3-D or higher with shape `[..., r_x, c_x]`
and `[..., r_y, c_y]`.</p><p>The output tensor is 3-D or higher with shape `[..., r_o, c_o]`, where:</p><p>r_o = c_x if adj_x else r_x
c_o = r_y if adj_y else c_y</p><p>It is computed as:</p><p>output[..., :, :] = matrix(x[..., :, :]) * matrix(y[..., :, :])</p></div></div><divclass="top"><pclass="src"><aname="v:refSelect"class="def">refSelect</a><ahref="src/TensorFlow-GenOps-Core.html#refSelect"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a></td><tdclass="doc"><p><strong>index</strong>: A scalar that determines the input that gets selected.</p></td></tr><tr><tdclass="src">-> [<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t]</td><tdclass="doc"><p><strong>inputs</strong>: A list of ref tensors, one of which will be forwarded to <code>output</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output</strong>: The forwarded tensor.</p></td></tr></table></div><divclass="doc"><p>Forwards the <code>index</code>th element of <code>inputs</code> to <code>output</code>.</p></div></div><divclass="top"><pclass="src"><aname="v:mean"class="def">mean</a><ahref="src/TensorFlow-GenOps-Core.html#mean"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tidx, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tidx)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>input</strong>: The tensor to reduce.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tidx</td><tdclass="doc"><p><strong>reduction_indices</strong>: The dimensions to reduce.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: The reduced tensor.</p></td></tr></table></div><divclass="doc"><p>Computes the mean of elements across dimensions of a tensor.</p><p>Reduces <code>input</code> along the dimensions given in <code>reduction_indices</code>. Unless
<code>keep_dims</code> is true, the rank of the tensor is reduced by 1 for each entry in
<code>reduction_indices</code>. If <code>keep_dims</code> is true, the reduced dimensions are
retained with length 1.</p></div></div><divclass="top"><pclass="src"><aname="v:scatterAdd"class="def">scatterAdd</a><ahref="src/TensorFlow-GenOps-Core.html#scatterAdd"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tindices, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` tindices)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>ref</strong>: Should be from a <code>Variable</code> node.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tindices</td><tdclass="doc"><p><strong>indices</strong>: A tensor of indices into the first dimension of <code>ref</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>updates</strong>: A tensor of updated values to add to <code>ref</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output_ref</strong>: = Same as <code>ref</code>. Returned as a convenience for operations that want
to use the updated values after the update is done.</p></td></tr></table></div><divclass="doc"><p>Adds sparse updates to a variable reference.</p><p>This operation computes</p><p># Scalar indices
ref[indices, ...] += updates[...]</p><p># Vector indices (for each i)
ref[indices[i], ...] += updates[i, ...]</p><p># High rank indices (for each i, ..., j)
ref[indices[i, ..., j], ...] += updates[i, ..., j, ...]</p><p>This operation outputs <code>ref</code> after the update is done.
This makes it easier to chain operations that need to use the reset value.</p><p>Duplicate entries are handled correctly: if multiple <code>indices</code> reference
the same location, their contributions add.</p><p>Requires `updates.shape = indices.shape + ref.shape[1:]`.</p><p><ahref="div">style="width:70%; margin:auto; margin-bottom:10px; margin-top:20px;"</a>
<ahref="/div">/div</a></p></div></div><divclass="top"><pclass="src"><aname="v:randomCrop"class="def">randomCrop</a><ahref="src/TensorFlow-GenOps-Core.html#randomCrop"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>image</strong>: 3-D of shape `[height, width, channels]`.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>size</strong>: 1-D of length 2 containing: <code>crop_height</code>, <code>crop_width</code>..</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p><strong>output</strong>: 3-D of shape `[crop_height, crop_width, channels].`</p></td></tr></table></div><divclass="doc"><p>Randomly crop <code>image</code>.</p><p><code><ahref="TensorFlow-GenOps-Core.html#v:size">size</a></code> is a 1-D int64 tensor with 2 elements representing the crop height and
width. The values must be non negative.</p><p>This Op picks a random location in <code>image</code> and crops a <code>height</code> by <code>width</code>
rectangle from that location. The random location is picked so the cropped
area will fit inside the original image.</p></div></div><divclass="top"><pclass="src"><aname="v:refExit"class="def">refExit</a><ahref="src/TensorFlow-GenOps-Core.html#refExit"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t</td><tdclass="doc"><p><strong>data</strong>: The tensor to be made available to the parent frame.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a> t)</td><tdclass="doc"><p><strong>output</strong>: The same tensor as `data`.</p></td></tr></table></div><divclass="doc"><p>Exits the current frame to its parent frame.</p><p>Exit makes its input `data` available to the parent frame.</p></div></div><divclass="top"><pclass="src"><aname="v:readerSerializeState"class="def">readerSerializeState</a><ahref="src/TensorFlow-GenOps-Core.html#readerSerializeState"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>reader_handle</strong>: Handle to a Reader.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a>)</td><tdclass="doc"><p><strong>state</strong></p></td></tr></table></div><divclass="doc"><p>Produce a string tensor that encodes the state of a Reader.</p><p>Not all Readers support being serialized, so this can produce an
Unimplemented error.</p></div></div><divclass="top"><pclass="src"><aname="v:tanhGrad"class="def">tanhGrad</a><ahref="src/TensorFlow-GenOps-Core.html#tanhGrad"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 t</td><tdclass="doc"><p><strong>x</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>y</strong></p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>z</strong></p></td></tr></table></div><divclass="doc"><p>Computes the gradient for the tanh of <code>x</code> wrt its input.</p><p>Specifically, `grad = dy * (1 - y*y)`, where `y = tanh(x)`, and <code>dy</code>
is the corresponding input gradient.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseSparseMaximum"class="def">sparseSparseMaximum</a><ahref="src/TensorFlow-GenOps-Core.html#sparseSparseMaximum"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>a_indices</strong>: 2-D. `N x R` matrix with the indices of non-empty values in a
SparseTensor, in the canonical lexicographic ordering.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>a_values</strong>: 1-D. <code>N</code> non-empty values corresponding to <code>a_indices</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>a_shape</strong>: 1-D. Shape of the input SparseTensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>b_indices</strong>: counterpart to <code>a_indices</code> for the other operand.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v5 t</td><tdclass="doc"><p><strong>b_values</strong>: counterpart to <code>a_values</code> for the other operand; must be of the same dtype.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v6 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>b_shape</strong>: counterpart to <code>a_shape</code> for the other operand; the two shapes must be equal.</p></td></tr><tr><tdclass="src">-> (<ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t)</td><tdclass="doc"><p>(<strong>output_indices</strong>, <strong>output_values</strong>)</p><ul><li><strong>output_indices</strong>: 2-D. The indices of the output SparseTensor.</li><li><strong>output_values</strong>: 1-D. The values of the output SparseTensor.</li></ul></td></tr></table></div><divclass="doc"><p>Returns the element-wise max of two SparseTensors.</p><p>Assumes the two SparseTensors have the same shape, i.e., no broadcasting.</p></div></div><divclass="top"><pclass="src"><aname="v:decodeGif"class="def">decodeGif</a><ahref="src/TensorFlow-GenOps-Core.html#decodeGif"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: <ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>contents</strong>: 0-D. The GIF-encoded image.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a></td><tdclass="doc"><p><strong>image</strong>: 4-D with shape `[num_frames, height, width, 3]`. RGB order</p></td></tr></table></div><divclass="doc"><p>Decode the first frame of a GIF-encoded image to a uint8 tensor.</p><p>GIF with frame or transparency compression are not supported
convert animated GIF from compressed to uncompressed by:</p><p>convert $src.gif -coalesce $dst.gif</p></div></div><divclass="top"><pclass="src"><aname="v:substr"class="def">substr</a><ahref="src/TensorFlow-GenOps-Core.html#substr"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>input</strong>: Tensor of strings</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>pos</strong>: Scalar defining the position of first character in each substring</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 t</td><tdclass="doc"><p><strong>len</strong>: Scalar defining the number of characters to include in each substring</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>output</strong>: Tensor of substrings</p></td></tr></table></div><divclass="doc"><p>Return substrings from <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code> of strings.</p><p>For each string in the input <code><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a></code>, creates a substring starting at index
<code>pos</code> with a total length of <code>len</code>.</p><p>If <code>len</code> defines a substring that would extend beyond the length of the input
string, then as many characters as possible are used.</p><p>If <code>pos</code> is negative or specifies a character index larger than any of the input
strings, then an <code>InvalidArgumentError</code> is thrown.</p><p><code>pos</code> and <code>len</code> must have the same shape, otherwise a <code>ValueError</code> is thrown on
Op creation.</p><ul><li>NOTE*: <code>Substr</code> supports broadcasting up to two dimensions. More about
broadcasting
<ahref="http://docs.scipy.org/doc/numpy/user/basics.broadcasting.html">here</a></li><li>--</li></ul><p>Examples</p><p>Using scalar <code>pos</code> and <code>len</code>:</p><p>```
```</p></div></div><divclass="top"><pclass="src"><aname="v:lookupTableInsert"class="def">lookupTableInsert</a><ahref="src/TensorFlow-GenOps-Core.html#lookupTableInsert"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tin, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tout)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>table_handle</strong>: Handle to the table.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tin</td><tdclass="doc"><p><strong>keys</strong>: Any shape. Keys to look up.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 tout</td><tdclass="doc"><p><strong>values</strong>: Values to associate with keys.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Updates the table to associates keys with values.</p><p>The tensor <code>keys</code> must be of the same type as the keys of the table.
The tensor <code>values</code> must be of the type of the table values.</p></div></div><divclass="top"><pclass="src"><aname="v:sparseDenseCwiseDiv"class="def">sparseDenseCwiseDiv</a><ahref="src/TensorFlow-GenOps-Core.html#sparseDenseCwiseDiv"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> t, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:OneOf">OneOf</a> `[<ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Data-Complex.html#t:Complex">Complex</a><ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int16">Int16</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int32">Int32</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a>, <ahref="../base-4.8.2.0/Data-Int.html#t:Int8">Int8</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word16">Word16</a>, <ahref="../base-4.8.2.0/Data-Word.html#t:Word8">Word8</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Double">Double</a>, <ahref="../base-4.8.2.0/Prelude.html#t:Float">Float</a>]` t)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v1 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sp_indices</strong>: 2-D. `N x R` matrix with the indices of non-empty values in a
SparseTensor, possibly not in canonical ordering.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 t</td><tdclass="doc"><p><strong>sp_values</strong>: 1-D. <code>N</code> non-empty values corresponding to <code>sp_indices</code>.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 <ahref="../base-4.8.2.0/Data-Int.html#t:Int64">Int64</a></td><tdclass="doc"><p><strong>sp_shape</strong>: 1-D. Shape of the input SparseTensor.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v4 t</td><tdclass="doc"><p><strong>dense</strong>: <code>R</code>-D. The dense Tensor operand.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Value">Value</a> t</td><tdclass="doc"><p><strong>output</strong>: 1-D. The <code>N</code> values that are operated on.</p></td></tr></table></div><divclass="doc"><p>Component-wise divides a SparseTensor by a dense Tensor.</p><ul><li>Limitation*: this Op only broadcasts the dense side to the sparse side, but not
the other direction.</li></ul></div></div><divclass="top"><pclass="src"><aname="v:lookupTableImport"class="def">lookupTableImport</a><ahref="src/TensorFlow-GenOps-Core.html#lookupTableImport"class="link">Source</a></p><divclass="subs arguments"><pclass="caption">Arguments</p><table><tr><tdclass="src">:: (<ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tin, <ahref="../tensorflow-0.1.0.0/TensorFlow-Types.html#t:TensorType">TensorType</a> tout)</td><tdclass="doc empty"> </td></tr><tr><tdclass="src">=><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Ref">Ref</a><ahref="../bytestring-0.10.6.0/Data-ByteString.html#t:ByteString">ByteString</a></td><tdclass="doc"><p><strong>table_handle</strong>: Handle to the table.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v2 tin</td><tdclass="doc"><p><strong>keys</strong>: Any shape. Keys to look up.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Tensor.html#t:Tensor">Tensor</a> v3 tout</td><tdclass="doc"><p><strong>values</strong>: Values to associate with keys.</p></td></tr><tr><tdclass="src">-><ahref="../tensorflow-0.1.0.0/TensorFlow-Build.html#t:Build">Build</a><ahref="../tensorflow-0.1.0.0/TensorFlow-Output.html#t:ControlNode">ControlNode</a></td><tdclass="doc empty"> </td></tr></table></div><divclass="doc"><p>Replaces the contents of the table with the specified keys and values.</p><p>The tensor <code>keys</code> must be of the same type as the keys of the table.
The tensor <code>values</code> must be of the type of the table values.</p></div></div></div></div><divid="footer"><p>Produced by <ahref="http://www.haskell.org/haddock/">Haddock</a> version 2.16.1</p></div></body></html>