| <!-- HTML header for doxygen 1.8.15--> |
| <!-- Remember to use version doxygen 1.8.15 +--> |
| <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> |
| <html xmlns="http://www.w3.org/1999/xhtml"> |
| <head> |
| <meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/> |
| <meta http-equiv="X-UA-Compatible" content="IE=9"/> |
| <meta name="generator" content="Doxygen 1.8.15"/> |
| <meta name="robots" content="NOINDEX, NOFOLLOW" /> <!-- Prevent indexing by search engines --> |
| <title>Compute Library: CLLogits1DMaxShiftExpSumKernel Class Reference</title> |
| <link href="tabs.css" rel="stylesheet" type="text/css"/> |
| <script type="text/javascript" src="jquery.js"></script> |
| <script type="text/javascript" src="dynsections.js"></script> |
| <link href="navtree.css" rel="stylesheet" type="text/css"/> |
| <script type="text/javascript" src="resize.js"></script> |
| <script type="text/javascript" src="navtreedata.js"></script> |
| <script type="text/javascript" src="navtree.js"></script> |
| <script type="text/javascript"> |
| /* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&dn=gpl-2.0.txt GPL-v2 */ |
| $(document).ready(initResizable); |
| /* @license-end */</script> |
| <link href="search/search.css" rel="stylesheet" type="text/css"/> |
| <script type="text/javascript" src="search/searchdata.js"></script> |
| <script type="text/javascript" src="search/search.js"></script> |
| <script type="text/x-mathjax-config"> |
| MathJax.Hub.Config({ |
| extensions: ["tex2jax.js"], |
| jax: ["input/TeX","output/HTML-CSS"], |
| }); |
| </script><script type="text/javascript" async="async" src="http://cdn.mathjax.org/mathjax/latest/MathJax.js"></script> |
| <link href="doxygen.css" rel="stylesheet" type="text/css" /> |
| <link href="stylesheet.css" rel="stylesheet" type="text/css"/> |
| </head> |
| <body> |
| <div id="top"><!-- do not remove this div, it is closed by doxygen! --> |
| <div id="titlearea"> |
| <table cellspacing="0" cellpadding="0"> |
| <tbody> |
| <tr style="height: 56px;"> |
| <img alt="Compute Library" src="https://raw.githubusercontent.com/ARM-software/ComputeLibrary/gh-pages/ACL_logo.png" style="max-width: 100%;margin-top: 15px;margin-left: 10px"/> |
| <td style="padding-left: 0.5em;"> |
| <div id="projectname"> |
|  <span id="projectnumber">19.02</span> |
| </div> |
| </td> |
| </tr> |
| </tbody> |
| </table> |
| </div> |
| <!-- end header part --> |
| <!-- Generated by Doxygen 1.8.15 --> |
| <script type="text/javascript"> |
| /* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&dn=gpl-2.0.txt GPL-v2 */ |
| var searchBox = new SearchBox("searchBox", "search",false,'Search'); |
| /* @license-end */ |
| </script> |
| <script type="text/javascript" src="menudata.js"></script> |
| <script type="text/javascript" src="menu.js"></script> |
| <script type="text/javascript"> |
| /* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&dn=gpl-2.0.txt GPL-v2 */ |
| $(function() { |
| initMenu('',true,false,'search.php','Search'); |
| $(document).ready(function() { init_search(); }); |
| }); |
| /* @license-end */</script> |
| <div id="main-nav"></div> |
| </div><!-- top --> |
| <div id="side-nav" class="ui-resizable side-nav-resizable"> |
| <div id="nav-tree"> |
| <div id="nav-tree-contents"> |
| <div id="nav-sync" class="sync"></div> |
| </div> |
| </div> |
| <div id="splitbar" style="-moz-user-select:none;" |
| class="ui-resizable-handle"> |
| </div> |
| </div> |
| <script type="text/javascript"> |
| /* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&dn=gpl-2.0.txt GPL-v2 */ |
| $(document).ready(function(){initNavTree('classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml','');}); |
| /* @license-end */ |
| </script> |
| <div id="doc-content"> |
| <!-- window showing the filter options --> |
| <div id="MSearchSelectWindow" |
| onmouseover="return searchBox.OnSearchSelectShow()" |
| onmouseout="return searchBox.OnSearchSelectHide()" |
| onkeydown="return searchBox.OnSearchSelectKey(event)"> |
| </div> |
| |
| <!-- iframe showing the search results (closed by default) --> |
| <div id="MSearchResultsWindow"> |
| <iframe src="javascript:void(0)" frameborder="0" |
| name="MSearchResults" id="MSearchResults"> |
| </iframe> |
| </div> |
| |
| <div class="header"> |
| <div class="summary"> |
| <a href="#pub-types">Public Types</a> | |
| <a href="#pub-methods">Public Member Functions</a> | |
| <a href="#pub-static-methods">Static Public Member Functions</a> </div> |
| <div class="headertitle"> |
| <div class="title">CLLogits1DMaxShiftExpSumKernel Class Reference</div> </div> |
| </div><!--header--> |
| <div class="contents"> |
| |
| <p>Interface for max, shifting, exponentiating and summing the logits. |
| <a href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#details">More...</a></p> |
| |
| <p><code>#include <<a class="el" href="_c_l_softmax_layer_kernel_8h_source.xhtml">CLSoftmaxLayerKernel.h</a>></code></p> |
| <div class="dynheader"> |
| Collaboration diagram for CLLogits1DMaxShiftExpSumKernel:</div> |
| <div class="dyncontent"> |
| <div class="center"><iframe scrolling="no" frameborder="0" src="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel__coll__graph.svg" width="214" height="202"><p><b>This browser is not able to show SVG: try Firefox, Chrome, Safari, or Opera instead.</b></p></iframe> |
| </div> |
| <center><span class="legend">[<a target="top" href="graph_legend.xhtml">legend</a>]</span></center></div> |
| <table class="memberdecls"> |
| <tr class="heading"><td colspan="2"><h2 class="groupheader"><a name="pub-types"></a> |
| Public Types</h2></td></tr> |
| <tr class="memitem:a6474b575b0139e3e1973acc9179556b8"><td class="memItemLeft" align="right" valign="top">using </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a6474b575b0139e3e1973acc9179556b8">ParallelReductionInfo</a> = std::tuple< bool, unsigned int ></td></tr> |
| <tr class="memdesc:a6474b575b0139e3e1973acc9179556b8"><td class="mdescLeft"> </td><td class="mdescRight">Info for whether a parallel reduction will be run and the vector size of the execution. <a href="#a6474b575b0139e3e1973acc9179556b8">More...</a><br /></td></tr> |
| <tr class="separator:a6474b575b0139e3e1973acc9179556b8"><td class="memSeparator" colspan="2"> </td></tr> |
| </table><table class="memberdecls"> |
| <tr class="heading"><td colspan="2"><h2 class="groupheader"><a name="pub-methods"></a> |
| Public Member Functions</h2></td></tr> |
| <tr class="memitem:ad265f3317602210f39f9bfc76df0f214"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#ad265f3317602210f39f9bfc76df0f214">CLLogits1DMaxShiftExpSumKernel</a> ()</td></tr> |
| <tr class="memdesc:ad265f3317602210f39f9bfc76df0f214"><td class="mdescLeft"> </td><td class="mdescRight">Default constructor. <a href="#ad265f3317602210f39f9bfc76df0f214">More...</a><br /></td></tr> |
| <tr class="separator:ad265f3317602210f39f9bfc76df0f214"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a7b6bf49de8edacec797d40532e57aff5"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a7b6bf49de8edacec797d40532e57aff5">CLLogits1DMaxShiftExpSumKernel</a> (const <a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a> &)=delete</td></tr> |
| <tr class="memdesc:a7b6bf49de8edacec797d40532e57aff5"><td class="mdescLeft"> </td><td class="mdescRight">Prevent instances of this class from being copied (As this class contains pointers) <a href="#a7b6bf49de8edacec797d40532e57aff5">More...</a><br /></td></tr> |
| <tr class="separator:a7b6bf49de8edacec797d40532e57aff5"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:afd28bfde2ba6f90c161eb1f2678cb6f9"><td class="memItemLeft" align="right" valign="top"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a> & </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#afd28bfde2ba6f90c161eb1f2678cb6f9">operator=</a> (const <a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a> &)=delete</td></tr> |
| <tr class="memdesc:afd28bfde2ba6f90c161eb1f2678cb6f9"><td class="mdescLeft"> </td><td class="mdescRight">Prevent instances of this class from being copied (As this class contains pointers) <a href="#afd28bfde2ba6f90c161eb1f2678cb6f9">More...</a><br /></td></tr> |
| <tr class="separator:afd28bfde2ba6f90c161eb1f2678cb6f9"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:af0cb102c35523ed86cae30a6cbce8409"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#af0cb102c35523ed86cae30a6cbce8409">CLLogits1DMaxShiftExpSumKernel</a> (<a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a> &&)=default</td></tr> |
| <tr class="memdesc:af0cb102c35523ed86cae30a6cbce8409"><td class="mdescLeft"> </td><td class="mdescRight">Allow instances of this class to be moved. <a href="#af0cb102c35523ed86cae30a6cbce8409">More...</a><br /></td></tr> |
| <tr class="separator:af0cb102c35523ed86cae30a6cbce8409"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:ac7ea34435f64f3e3dce08ab0ce67295e"><td class="memItemLeft" align="right" valign="top"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a> & </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#ac7ea34435f64f3e3dce08ab0ce67295e">operator=</a> (<a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a> &&)=default</td></tr> |
| <tr class="memdesc:ac7ea34435f64f3e3dce08ab0ce67295e"><td class="mdescLeft"> </td><td class="mdescRight">Allow instances of this class to be moved. <a href="#ac7ea34435f64f3e3dce08ab0ce67295e">More...</a><br /></td></tr> |
| <tr class="separator:ac7ea34435f64f3e3dce08ab0ce67295e"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:ac168c44d948adca330a511f2c5c5664e"><td class="memItemLeft" align="right" valign="top">void </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#ac168c44d948adca330a511f2c5c5664e">configure</a> (const <a class="el" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *input, <a class="el" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *max, <a class="el" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *output, <a class="el" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *<a class="el" href="reduction__operation_8cl.xhtml#ab0df00f5333da51860deb93deb44a782">sum</a>, float beta=1.0f)</td></tr> |
| <tr class="memdesc:ac168c44d948adca330a511f2c5c5664e"><td class="mdescLeft"> </td><td class="mdescRight">Set the input and output tensors. <a href="#ac168c44d948adca330a511f2c5c5664e">More...</a><br /></td></tr> |
| <tr class="separator:ac168c44d948adca330a511f2c5c5664e"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a493987e85723a8000eb26d1f00e2ad0e"><td class="memItemLeft" align="right" valign="top">void </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a493987e85723a8000eb26d1f00e2ad0e">run</a> (const <a class="el" href="classarm__compute_1_1_window.xhtml">Window</a> &<a class="el" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a>, cl::CommandQueue &queue) override</td></tr> |
| <tr class="memdesc:a493987e85723a8000eb26d1f00e2ad0e"><td class="mdescLeft"> </td><td class="mdescRight">Enqueue the OpenCL kernel to process the given window on the passed OpenCL command queue. <a href="#a493987e85723a8000eb26d1f00e2ad0e">More...</a><br /></td></tr> |
| <tr class="separator:a493987e85723a8000eb26d1f00e2ad0e"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="inherit_header pub_methods_classarm__compute_1_1_i_c_l_kernel"><td colspan="2" onclick="javascript:toggleInherit('pub_methods_classarm__compute_1_1_i_c_l_kernel')"><img src="closed.png" alt="-"/> Public Member Functions inherited from <a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml">ICLKernel</a></td></tr> |
| <tr class="memitem:a6b10e96ce90bf901d17def86b874b019 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a6b10e96ce90bf901d17def86b874b019">ICLKernel</a> ()</td></tr> |
| <tr class="memdesc:a6b10e96ce90bf901d17def86b874b019 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Constructor. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a6b10e96ce90bf901d17def86b874b019">More...</a><br /></td></tr> |
| <tr class="separator:a6b10e96ce90bf901d17def86b874b019 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:ae5121015ab09ece4d470f50c7ffe198e inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">cl::Kernel & </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ae5121015ab09ece4d470f50c7ffe198e">kernel</a> ()</td></tr> |
| <tr class="memdesc:ae5121015ab09ece4d470f50c7ffe198e inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Returns a reference to the OpenCL kernel of this object. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#ae5121015ab09ece4d470f50c7ffe198e">More...</a><br /></td></tr> |
| <tr class="separator:ae5121015ab09ece4d470f50c7ffe198e inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a9331d385192a50adf74d3af40ce0fa20 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memTemplParams" colspan="2">template<typename T > </td></tr> |
| <tr class="memitem:a9331d385192a50adf74d3af40ce0fa20 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memTemplItemLeft" align="right" valign="top">void </td><td class="memTemplItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a9331d385192a50adf74d3af40ce0fa20">add_1D_array_argument</a> (unsigned int &idx, const <a class="el" href="classarm__compute_1_1_i_c_l_array.xhtml">ICLArray</a>< T > *array, const <a class="el" href="classarm__compute_1_1_strides.xhtml">Strides</a> &strides, unsigned int num_dimensions, const <a class="el" href="classarm__compute_1_1_window.xhtml">Window</a> &<a class="el" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a>)</td></tr> |
| <tr class="memdesc:a9331d385192a50adf74d3af40ce0fa20 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Add the passed 1D array's parameters to the object's kernel's arguments starting from the index idx. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a9331d385192a50adf74d3af40ce0fa20">More...</a><br /></td></tr> |
| <tr class="separator:a9331d385192a50adf74d3af40ce0fa20 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a479e7043e65dc87de35d374e108510f7 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">void </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a479e7043e65dc87de35d374e108510f7">add_1D_tensor_argument</a> (unsigned int &idx, const <a class="el" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *tensor, const <a class="el" href="classarm__compute_1_1_window.xhtml">Window</a> &<a class="el" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a>)</td></tr> |
| <tr class="memdesc:a479e7043e65dc87de35d374e108510f7 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Add the passed 1D tensor's parameters to the object's kernel's arguments starting from the index idx. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a479e7043e65dc87de35d374e108510f7">More...</a><br /></td></tr> |
| <tr class="separator:a479e7043e65dc87de35d374e108510f7 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:ac74dad3e61f79334f5e73f3c3ac603cb inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">void </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ac74dad3e61f79334f5e73f3c3ac603cb">add_2D_tensor_argument</a> (unsigned int &idx, const <a class="el" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *tensor, const <a class="el" href="classarm__compute_1_1_window.xhtml">Window</a> &<a class="el" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a>)</td></tr> |
| <tr class="memdesc:ac74dad3e61f79334f5e73f3c3ac603cb inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Add the passed 2D tensor's parameters to the object's kernel's arguments starting from the index idx. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#ac74dad3e61f79334f5e73f3c3ac603cb">More...</a><br /></td></tr> |
| <tr class="separator:ac74dad3e61f79334f5e73f3c3ac603cb inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a28f5847162f352444c6ac1825d0e99c7 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">void </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a28f5847162f352444c6ac1825d0e99c7">add_3D_tensor_argument</a> (unsigned int &idx, const <a class="el" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *tensor, const <a class="el" href="classarm__compute_1_1_window.xhtml">Window</a> &<a class="el" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a>)</td></tr> |
| <tr class="memdesc:a28f5847162f352444c6ac1825d0e99c7 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Add the passed 3D tensor's parameters to the object's kernel's arguments starting from the index idx. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a28f5847162f352444c6ac1825d0e99c7">More...</a><br /></td></tr> |
| <tr class="separator:a28f5847162f352444c6ac1825d0e99c7 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a33e09c946b338fbfc780a9d1c66e68ad inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">void </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a33e09c946b338fbfc780a9d1c66e68ad">add_4D_tensor_argument</a> (unsigned int &idx, const <a class="el" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *tensor, const <a class="el" href="classarm__compute_1_1_window.xhtml">Window</a> &<a class="el" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a>)</td></tr> |
| <tr class="memdesc:a33e09c946b338fbfc780a9d1c66e68ad inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Add the passed 4D tensor's parameters to the object's kernel's arguments starting from the index idx. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a33e09c946b338fbfc780a9d1c66e68ad">More...</a><br /></td></tr> |
| <tr class="separator:a33e09c946b338fbfc780a9d1c66e68ad inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a50f427a1d9419800972b9e03c4034311 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memTemplParams" colspan="2">template<typename T > </td></tr> |
| <tr class="memitem:a50f427a1d9419800972b9e03c4034311 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memTemplItemLeft" align="right" valign="top">void </td><td class="memTemplItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a50f427a1d9419800972b9e03c4034311">add_argument</a> (unsigned int &idx, T value)</td></tr> |
| <tr class="memdesc:a50f427a1d9419800972b9e03c4034311 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Add the passed parameters to the object's kernel's arguments starting from the index idx. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a50f427a1d9419800972b9e03c4034311">More...</a><br /></td></tr> |
| <tr class="separator:a50f427a1d9419800972b9e03c4034311 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:ad356b88c8f61267d593d9ed99835bde9 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">void </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ad356b88c8f61267d593d9ed99835bde9">set_lws_hint</a> (const cl::NDRange &<a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ab9f813c25ed75ea7b7ac2fa3926a8f55">lws_hint</a>)</td></tr> |
| <tr class="memdesc:ad356b88c8f61267d593d9ed99835bde9 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Set the Local-Workgroup-Size hint. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#ad356b88c8f61267d593d9ed99835bde9">More...</a><br /></td></tr> |
| <tr class="separator:ad356b88c8f61267d593d9ed99835bde9 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:ab9f813c25ed75ea7b7ac2fa3926a8f55 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">cl::NDRange </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ab9f813c25ed75ea7b7ac2fa3926a8f55">lws_hint</a> () const</td></tr> |
| <tr class="memdesc:ab9f813c25ed75ea7b7ac2fa3926a8f55 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Return the Local-Workgroup-Size hint. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#ab9f813c25ed75ea7b7ac2fa3926a8f55">More...</a><br /></td></tr> |
| <tr class="separator:ab9f813c25ed75ea7b7ac2fa3926a8f55 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a8f7f6ab59fc0e601d750b83e75a398eb inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">const std::string & </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a8f7f6ab59fc0e601d750b83e75a398eb">config_id</a> () const</td></tr> |
| <tr class="memdesc:a8f7f6ab59fc0e601d750b83e75a398eb inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Get the configuration ID. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a8f7f6ab59fc0e601d750b83e75a398eb">More...</a><br /></td></tr> |
| <tr class="separator:a8f7f6ab59fc0e601d750b83e75a398eb inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:ad5ba9d34a3a855bf1dd2e36316ff550a inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">void </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ad5ba9d34a3a855bf1dd2e36316ff550a">set_target</a> (<a class="el" href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3">GPUTarget</a> target)</td></tr> |
| <tr class="memdesc:ad5ba9d34a3a855bf1dd2e36316ff550a inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Set the targeted GPU architecture. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#ad5ba9d34a3a855bf1dd2e36316ff550a">More...</a><br /></td></tr> |
| <tr class="separator:ad5ba9d34a3a855bf1dd2e36316ff550a inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a57e5f498fcbfc25c28b8496dfa3fc33c inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">void </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a57e5f498fcbfc25c28b8496dfa3fc33c">set_target</a> (cl::Device &device)</td></tr> |
| <tr class="memdesc:a57e5f498fcbfc25c28b8496dfa3fc33c inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Set the targeted GPU architecture according to the CL device. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a57e5f498fcbfc25c28b8496dfa3fc33c">More...</a><br /></td></tr> |
| <tr class="separator:a57e5f498fcbfc25c28b8496dfa3fc33c inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:aa550ff0352ff2388e02f7b0a41bf5fe7 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top"><a class="el" href="namespacearm__compute.xhtml#a735ac6c2a02e320969625308810444f3">GPUTarget</a> </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#aa550ff0352ff2388e02f7b0a41bf5fe7">get_target</a> () const</td></tr> |
| <tr class="memdesc:aa550ff0352ff2388e02f7b0a41bf5fe7 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Get the targeted GPU architecture. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#aa550ff0352ff2388e02f7b0a41bf5fe7">More...</a><br /></td></tr> |
| <tr class="separator:aa550ff0352ff2388e02f7b0a41bf5fe7 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:abca336f832d730e8494049bd714df60a inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">size_t </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#abca336f832d730e8494049bd714df60a">get_max_workgroup_size</a> ()</td></tr> |
| <tr class="memdesc:abca336f832d730e8494049bd714df60a inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Get the maximum workgroup size for the device the <a class="el" href="classarm__compute_1_1_c_l_kernel_library.xhtml" title="CLKernelLibrary class.">CLKernelLibrary</a> uses. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#abca336f832d730e8494049bd714df60a">More...</a><br /></td></tr> |
| <tr class="separator:abca336f832d730e8494049bd714df60a inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a2d7c6b5f3332604ad6a637457f65c342 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memTemplParams" colspan="2">template<typename T , unsigned int dimension_size> </td></tr> |
| <tr class="memitem:a2d7c6b5f3332604ad6a637457f65c342 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memTemplItemLeft" align="right" valign="top">void </td><td class="memTemplItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a2d7c6b5f3332604ad6a637457f65c342">add_array_argument</a> (unsigned &idx, const <a class="el" href="classarm__compute_1_1_i_c_l_array.xhtml">ICLArray</a>< T > *array, const <a class="el" href="classarm__compute_1_1_strides.xhtml">Strides</a> &strides, unsigned int num_dimensions, const <a class="el" href="classarm__compute_1_1_window.xhtml">Window</a> &<a class="el" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a>)</td></tr> |
| <tr class="memdesc:a2d7c6b5f3332604ad6a637457f65c342 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Add the passed array's parameters to the object's kernel's arguments starting from the index idx. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a2d7c6b5f3332604ad6a637457f65c342">More...</a><br /></td></tr> |
| <tr class="separator:a2d7c6b5f3332604ad6a637457f65c342 inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:aa0191633c258a30cc27e35b99b663aac inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memTemplParams" colspan="2">template<unsigned int dimension_size> </td></tr> |
| <tr class="memitem:aa0191633c258a30cc27e35b99b663aac inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memTemplItemLeft" align="right" valign="top">void </td><td class="memTemplItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#aa0191633c258a30cc27e35b99b663aac">add_tensor_argument</a> (unsigned &idx, const <a class="el" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *tensor, const <a class="el" href="classarm__compute_1_1_window.xhtml">Window</a> &<a class="el" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a>)</td></tr> |
| <tr class="separator:aa0191633c258a30cc27e35b99b663aac inherit pub_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="inherit_header pub_methods_classarm__compute_1_1_i_kernel"><td colspan="2" onclick="javascript:toggleInherit('pub_methods_classarm__compute_1_1_i_kernel')"><img src="closed.png" alt="-"/> Public Member Functions inherited from <a class="el" href="classarm__compute_1_1_i_kernel.xhtml">IKernel</a></td></tr> |
| <tr class="memitem:a7250cb8cbaa4104a93a2d77155085507 inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="memItemLeft" align="right" valign="top"> </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_kernel.xhtml#a7250cb8cbaa4104a93a2d77155085507">IKernel</a> ()</td></tr> |
| <tr class="memdesc:a7250cb8cbaa4104a93a2d77155085507 inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Constructor. <a href="classarm__compute_1_1_i_kernel.xhtml#a7250cb8cbaa4104a93a2d77155085507">More...</a><br /></td></tr> |
| <tr class="separator:a7250cb8cbaa4104a93a2d77155085507 inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a341b60d15a5e12a5b8f3825194dd3b12 inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="memItemLeft" align="right" valign="top">virtual </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_kernel.xhtml#a341b60d15a5e12a5b8f3825194dd3b12">~IKernel</a> ()=default</td></tr> |
| <tr class="memdesc:a341b60d15a5e12a5b8f3825194dd3b12 inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Destructor. <a href="classarm__compute_1_1_i_kernel.xhtml#a341b60d15a5e12a5b8f3825194dd3b12">More...</a><br /></td></tr> |
| <tr class="separator:a341b60d15a5e12a5b8f3825194dd3b12 inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a0466ee6ce6552c87595f0e88e73eeb1b inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="memItemLeft" align="right" valign="top">virtual bool </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_kernel.xhtml#a0466ee6ce6552c87595f0e88e73eeb1b">is_parallelisable</a> () const</td></tr> |
| <tr class="memdesc:a0466ee6ce6552c87595f0e88e73eeb1b inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Indicates whether or not the kernel is parallelisable. <a href="classarm__compute_1_1_i_kernel.xhtml#a0466ee6ce6552c87595f0e88e73eeb1b">More...</a><br /></td></tr> |
| <tr class="separator:a0466ee6ce6552c87595f0e88e73eeb1b inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a4b3a97ba5dded504a2f2261c078493dd inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="memItemLeft" align="right" valign="top">virtual <a class="el" href="structarm__compute_1_1_border_size.xhtml">BorderSize</a> </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_kernel.xhtml#a4b3a97ba5dded504a2f2261c078493dd">border_size</a> () const</td></tr> |
| <tr class="memdesc:a4b3a97ba5dded504a2f2261c078493dd inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="mdescLeft"> </td><td class="mdescRight">The size of the border for that kernel. <a href="classarm__compute_1_1_i_kernel.xhtml#a4b3a97ba5dded504a2f2261c078493dd">More...</a><br /></td></tr> |
| <tr class="separator:a4b3a97ba5dded504a2f2261c078493dd inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:ad34a46f53686c12a5c5e717cc9617fb6 inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="memItemLeft" align="right" valign="top">const <a class="el" href="classarm__compute_1_1_window.xhtml">Window</a> & </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a> () const</td></tr> |
| <tr class="memdesc:ad34a46f53686c12a5c5e717cc9617fb6 inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="mdescLeft"> </td><td class="mdescRight">The maximum window the kernel can be executed on. <a href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">More...</a><br /></td></tr> |
| <tr class="separator:ad34a46f53686c12a5c5e717cc9617fb6 inherit pub_methods_classarm__compute_1_1_i_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| </table><table class="memberdecls"> |
| <tr class="heading"><td colspan="2"><h2 class="groupheader"><a name="pub-static-methods"></a> |
| Static Public Member Functions</h2></td></tr> |
| <tr class="memitem:a62885f3bde4b458c02316bfc3f339fbc"><td class="memItemLeft" align="right" valign="top">static <a class="el" href="classarm__compute_1_1_status.xhtml">Status</a> </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a62885f3bde4b458c02316bfc3f339fbc">validate</a> (const <a class="el" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *input, const <a class="el" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *max, const <a class="el" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *output, const <a class="el" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *<a class="el" href="reduction__operation_8cl.xhtml#ab0df00f5333da51860deb93deb44a782">sum</a>)</td></tr> |
| <tr class="memdesc:a62885f3bde4b458c02316bfc3f339fbc"><td class="mdescLeft"> </td><td class="mdescRight">Static function to check if given info will lead to a valid configuration of <a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a>. <a href="#a62885f3bde4b458c02316bfc3f339fbc">More...</a><br /></td></tr> |
| <tr class="separator:a62885f3bde4b458c02316bfc3f339fbc"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a78a33824fc1fd7c44535530ebd917eb4"><td class="memItemLeft" align="right" valign="top">static <a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a6474b575b0139e3e1973acc9179556b8">ParallelReductionInfo</a> </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a78a33824fc1fd7c44535530ebd917eb4">is_parallel_reduction</a> (size_t size)</td></tr> |
| <tr class="memdesc:a78a33824fc1fd7c44535530ebd917eb4"><td class="mdescLeft"> </td><td class="mdescRight">Checks if the given size is eligible for parallel reduction. <a href="#a78a33824fc1fd7c44535530ebd917eb4">More...</a><br /></td></tr> |
| <tr class="separator:a78a33824fc1fd7c44535530ebd917eb4"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="inherit_header pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td colspan="2" onclick="javascript:toggleInherit('pub_static_methods_classarm__compute_1_1_i_c_l_kernel')"><img src="closed.png" alt="-"/> Static Public Member Functions inherited from <a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml">ICLKernel</a></td></tr> |
| <tr class="memitem:a278f0e6c68ca17e71f4c4ff82f360aa0 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">static constexpr unsigned int </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a278f0e6c68ca17e71f4c4ff82f360aa0">num_arguments_per_1D_array</a> ()</td></tr> |
| <tr class="memdesc:a278f0e6c68ca17e71f4c4ff82f360aa0 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Returns the number of arguments enqueued per 1D array object. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a278f0e6c68ca17e71f4c4ff82f360aa0">More...</a><br /></td></tr> |
| <tr class="separator:a278f0e6c68ca17e71f4c4ff82f360aa0 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a43b6c5e4b57069c5f61e96dff24c212d inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">static constexpr unsigned int </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a43b6c5e4b57069c5f61e96dff24c212d">num_arguments_per_1D_tensor</a> ()</td></tr> |
| <tr class="memdesc:a43b6c5e4b57069c5f61e96dff24c212d inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Returns the number of arguments enqueued per 1D tensor object. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a43b6c5e4b57069c5f61e96dff24c212d">More...</a><br /></td></tr> |
| <tr class="separator:a43b6c5e4b57069c5f61e96dff24c212d inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a45601e0d46621a5b6f2e417d60e5c800 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">static constexpr unsigned int </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a45601e0d46621a5b6f2e417d60e5c800">num_arguments_per_2D_tensor</a> ()</td></tr> |
| <tr class="memdesc:a45601e0d46621a5b6f2e417d60e5c800 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Returns the number of arguments enqueued per 2D tensor object. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a45601e0d46621a5b6f2e417d60e5c800">More...</a><br /></td></tr> |
| <tr class="separator:a45601e0d46621a5b6f2e417d60e5c800 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a6c9c1e7a7d96743375ca40847f0f12e2 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">static constexpr unsigned int </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a6c9c1e7a7d96743375ca40847f0f12e2">num_arguments_per_3D_tensor</a> ()</td></tr> |
| <tr class="memdesc:a6c9c1e7a7d96743375ca40847f0f12e2 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Returns the number of arguments enqueued per 3D tensor object. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a6c9c1e7a7d96743375ca40847f0f12e2">More...</a><br /></td></tr> |
| <tr class="separator:a6c9c1e7a7d96743375ca40847f0f12e2 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a184fdf37587a9314cf12623accea6c73 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">static constexpr unsigned int </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a184fdf37587a9314cf12623accea6c73">num_arguments_per_4D_tensor</a> ()</td></tr> |
| <tr class="memdesc:a184fdf37587a9314cf12623accea6c73 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Returns the number of arguments enqueued per 4D tensor object. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a184fdf37587a9314cf12623accea6c73">More...</a><br /></td></tr> |
| <tr class="separator:a184fdf37587a9314cf12623accea6c73 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| <tr class="memitem:a6c01790e4e3f22f70f69002f0cb1b913 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memItemLeft" align="right" valign="top">static cl::NDRange </td><td class="memItemRight" valign="bottom"><a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a6c01790e4e3f22f70f69002f0cb1b913">gws_from_window</a> (const <a class="el" href="classarm__compute_1_1_window.xhtml">Window</a> &<a class="el" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a>)</td></tr> |
| <tr class="memdesc:a6c01790e4e3f22f70f69002f0cb1b913 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="mdescLeft"> </td><td class="mdescRight">Get the global work size given an execution window. <a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a6c01790e4e3f22f70f69002f0cb1b913">More...</a><br /></td></tr> |
| <tr class="separator:a6c01790e4e3f22f70f69002f0cb1b913 inherit pub_static_methods_classarm__compute_1_1_i_c_l_kernel"><td class="memSeparator" colspan="2"> </td></tr> |
| </table> |
| <a name="details" id="details"></a><h2 class="groupheader">Detailed Description</h2> |
| <div class="textblock"><p>Interface for max, shifting, exponentiating and summing the logits. </p> |
| |
| <p class="definition">Definition at line <a class="el" href="_c_l_softmax_layer_kernel_8h_source.xhtml#l00100">100</a> of file <a class="el" href="_c_l_softmax_layer_kernel_8h_source.xhtml">CLSoftmaxLayerKernel.h</a>.</p> |
| </div><h2 class="groupheader">Member Typedef Documentation</h2> |
| <a id="a6474b575b0139e3e1973acc9179556b8"></a> |
| <h2 class="memtitle"><span class="permalink"><a href="#a6474b575b0139e3e1973acc9179556b8">◆ </a></span>ParallelReductionInfo</h2> |
| |
| <div class="memitem"> |
| <div class="memproto"> |
| <table class="memname"> |
| <tr> |
| <td class="memname">using <a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a6474b575b0139e3e1973acc9179556b8">ParallelReductionInfo</a> = std::tuple<bool, unsigned int></td> |
| </tr> |
| </table> |
| </div><div class="memdoc"> |
| |
| <p>Info for whether a parallel reduction will be run and the vector size of the execution. </p> |
| |
| <p class="definition">Definition at line <a class="el" href="_c_l_softmax_layer_kernel_8h_source.xhtml#l00104">104</a> of file <a class="el" href="_c_l_softmax_layer_kernel_8h_source.xhtml">CLSoftmaxLayerKernel.h</a>.</p> |
| |
| </div> |
| </div> |
| <h2 class="groupheader">Constructor & Destructor Documentation</h2> |
| <a id="ad265f3317602210f39f9bfc76df0f214"></a> |
| <h2 class="memtitle"><span class="permalink"><a href="#ad265f3317602210f39f9bfc76df0f214">◆ </a></span>CLLogits1DMaxShiftExpSumKernel() <span class="overload">[1/3]</span></h2> |
| |
| <div class="memitem"> |
| <div class="memproto"> |
| <table class="memname"> |
| <tr> |
| <td class="memname"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a> </td> |
| <td>(</td> |
| <td class="paramname"></td><td>)</td> |
| <td></td> |
| </tr> |
| </table> |
| </div><div class="memdoc"> |
| |
| <p>Default constructor. </p> |
| |
| <p class="definition">Definition at line <a class="el" href="_c_l_softmax_layer_kernel_8cpp_source.xhtml#l00214">214</a> of file <a class="el" href="_c_l_softmax_layer_kernel_8cpp_source.xhtml">CLSoftmaxLayerKernel.cpp</a>.</p> |
| <div class="fragment"><div class="line"><a name="l00215"></a><span class="lineno"> 215</span>  : _input(<span class="keyword">nullptr</span>), _max(<span class="keyword">nullptr</span>), _output(<span class="keyword">nullptr</span>), _sum(<span class="keyword">nullptr</span>)</div><div class="line"><a name="l00216"></a><span class="lineno"> 216</span> {</div><div class="line"><a name="l00217"></a><span class="lineno"> 217</span> }</div></div><!-- fragment --> |
| </div> |
| </div> |
| <a id="a7b6bf49de8edacec797d40532e57aff5"></a> |
| <h2 class="memtitle"><span class="permalink"><a href="#a7b6bf49de8edacec797d40532e57aff5">◆ </a></span>CLLogits1DMaxShiftExpSumKernel() <span class="overload">[2/3]</span></h2> |
| |
| <div class="memitem"> |
| <div class="memproto"> |
| <table class="mlabels"> |
| <tr> |
| <td class="mlabels-left"> |
| <table class="memname"> |
| <tr> |
| <td class="memname"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a> </td> |
| <td>(</td> |
| <td class="paramtype">const <a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a> & </td> |
| <td class="paramname"></td><td>)</td> |
| <td></td> |
| </tr> |
| </table> |
| </td> |
| <td class="mlabels-right"> |
| <span class="mlabels"><span class="mlabel">delete</span></span> </td> |
| </tr> |
| </table> |
| </div><div class="memdoc"> |
| |
| <p>Prevent instances of this class from being copied (As this class contains pointers) </p> |
| |
| </div> |
| </div> |
| <a id="af0cb102c35523ed86cae30a6cbce8409"></a> |
| <h2 class="memtitle"><span class="permalink"><a href="#af0cb102c35523ed86cae30a6cbce8409">◆ </a></span>CLLogits1DMaxShiftExpSumKernel() <span class="overload">[3/3]</span></h2> |
| |
| <div class="memitem"> |
| <div class="memproto"> |
| <table class="mlabels"> |
| <tr> |
| <td class="mlabels-left"> |
| <table class="memname"> |
| <tr> |
| <td class="memname"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a> </td> |
| <td>(</td> |
| <td class="paramtype"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a> && </td> |
| <td class="paramname"></td><td>)</td> |
| <td></td> |
| </tr> |
| </table> |
| </td> |
| <td class="mlabels-right"> |
| <span class="mlabels"><span class="mlabel">default</span></span> </td> |
| </tr> |
| </table> |
| </div><div class="memdoc"> |
| |
| <p>Allow instances of this class to be moved. </p> |
| |
| </div> |
| </div> |
| <h2 class="groupheader">Member Function Documentation</h2> |
| <a id="ac168c44d948adca330a511f2c5c5664e"></a> |
| <h2 class="memtitle"><span class="permalink"><a href="#ac168c44d948adca330a511f2c5c5664e">◆ </a></span>configure()</h2> |
| |
| <div class="memitem"> |
| <div class="memproto"> |
| <table class="memname"> |
| <tr> |
| <td class="memname">void configure </td> |
| <td>(</td> |
| <td class="paramtype">const <a class="el" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> * </td> |
| <td class="paramname"><em>input</em>, </td> |
| </tr> |
| <tr> |
| <td class="paramkey"></td> |
| <td></td> |
| <td class="paramtype"><a class="el" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> * </td> |
| <td class="paramname"><em>max</em>, </td> |
| </tr> |
| <tr> |
| <td class="paramkey"></td> |
| <td></td> |
| <td class="paramtype"><a class="el" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> * </td> |
| <td class="paramname"><em>output</em>, </td> |
| </tr> |
| <tr> |
| <td class="paramkey"></td> |
| <td></td> |
| <td class="paramtype"><a class="el" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> * </td> |
| <td class="paramname"><em>sum</em>, </td> |
| </tr> |
| <tr> |
| <td class="paramkey"></td> |
| <td></td> |
| <td class="paramtype">float </td> |
| <td class="paramname"><em>beta</em> = <code>1.0f</code> </td> |
| </tr> |
| <tr> |
| <td></td> |
| <td>)</td> |
| <td></td><td></td> |
| </tr> |
| </table> |
| </div><div class="memdoc"> |
| |
| <p>Set the input and output tensors. </p> |
| <dl class="params"><dt>Parameters</dt><dd> |
| <table class="params"> |
| <tr><td class="paramdir">[in]</td><td class="paramname">input</td><td>Source tensor. Data types supported: F16/F32 </td></tr> |
| <tr><td class="paramdir">[in,out]</td><td class="paramname">max</td><td>Max values tensor. Data types supported: same as <code>input</code> </td></tr> |
| <tr><td class="paramdir">[out]</td><td class="paramname">output</td><td>Destination tensor. Data types supported: same as <code>input</code> </td></tr> |
| <tr><td class="paramdir">[out]</td><td class="paramname">sum</td><td>Sum of 1D logits tensor. Data types supported: same as <code>input</code> </td></tr> |
| <tr><td class="paramdir">[in]</td><td class="paramname">beta</td><td>(Optional) A scaling factor for the exponent. Defaults to 1.f </td></tr> |
| </table> |
| </dd> |
| </dl> |
| |
| <p class="definition">Definition at line <a class="el" href="_c_l_softmax_layer_kernel_8cpp_source.xhtml#l00219">219</a> of file <a class="el" href="_c_l_softmax_layer_kernel_8cpp_source.xhtml">CLSoftmaxLayerKernel.cpp</a>.</p> |
| <div class="fragment"><div class="line"><a name="l00220"></a><span class="lineno"> 220</span> {</div><div class="line"><a name="l00221"></a><span class="lineno"> 221</span>  <a class="code" href="_validate_8h.xhtml#a921b705e9e3e0fe928928447869e62a5">ARM_COMPUTE_ERROR_ON_NULLPTR</a>(input, max, <a class="code" href="reduction__operation_8cl.xhtml#ab0df00f5333da51860deb93deb44a782">sum</a>, output);</div><div class="line"><a name="l00222"></a><span class="lineno"> 222</span> </div><div class="line"><a name="l00223"></a><span class="lineno"> 223</span>  <span class="comment">// Output auto initialization if not yet initialized</span></div><div class="line"><a name="l00224"></a><span class="lineno"> 224</span>  <a class="code" href="namespacearm__compute.xhtml#a47be6fa38308d0003c25b60b7dbc45ce">auto_init_if_empty</a>(*<a class="code" href="reduction__operation_8cl.xhtml#ab0df00f5333da51860deb93deb44a782">sum</a>->info(), input-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">clone</a>()->set_tensor_shape(max-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a7c66505457d00ece3aa4b34cab80757d">tensor_shape</a>()));</div><div class="line"><a name="l00225"></a><span class="lineno"> 225</span>  <a class="code" href="namespacearm__compute.xhtml#a47be6fa38308d0003c25b60b7dbc45ce">auto_init_if_empty</a>(*output-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>(), *input-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">clone</a>());</div><div class="line"><a name="l00226"></a><span class="lineno"> 226</span> </div><div class="line"><a name="l00227"></a><span class="lineno"> 227</span>  <span class="comment">// Perform validation step</span></div><div class="line"><a name="l00228"></a><span class="lineno"> 228</span>  <a class="code" href="_error_8h.xhtml#a938dcd406ce611ef5345ad2531cdb948">ARM_COMPUTE_ERROR_THROW_ON</a>(validate_arguments_1DMaxShiftExpSum(input-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>(), max-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>(), output-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>(), <a class="code" href="reduction__operation_8cl.xhtml#ab0df00f5333da51860deb93deb44a782">sum</a>->info()));</div><div class="line"><a name="l00229"></a><span class="lineno"> 229</span> </div><div class="line"><a name="l00230"></a><span class="lineno"> 230</span>  _input = input;</div><div class="line"><a name="l00231"></a><span class="lineno"> 231</span>  _max = max;</div><div class="line"><a name="l00232"></a><span class="lineno"> 232</span>  _output = output;</div><div class="line"><a name="l00233"></a><span class="lineno"> 233</span>  _sum = <a class="code" href="reduction__operation_8cl.xhtml#ab0df00f5333da51860deb93deb44a782">sum</a>;</div><div class="line"><a name="l00234"></a><span class="lineno"> 234</span> </div><div class="line"><a name="l00235"></a><span class="lineno"> 235</span>  <span class="keyword">const</span> <a class="code" href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6">DataType</a> dt = input-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a7cfb31af63202568efef5214acfbf3ba">data_type</a>();</div><div class="line"><a name="l00236"></a><span class="lineno"> 236</span>  <span class="keyword">const</span> <span class="keywordtype">size_t</span> reduction_dim_size = input-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(0);</div><div class="line"><a name="l00237"></a><span class="lineno"> 237</span> </div><div class="line"><a name="l00238"></a><span class="lineno"> 238</span>  <span class="comment">// Set build options</span></div><div class="line"><a name="l00239"></a><span class="lineno"> 239</span>  <a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml">CLBuildOptions</a> build_opts;</div><div class="line"><a name="l00240"></a><span class="lineno"> 240</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a3e2b80ff5463b7d2017de847f5c32a30">add_option</a>(<span class="stringliteral">"-DDATA_TYPE="</span> + <a class="code" href="namespacearm__compute.xhtml#a545eeda2eaa3f5a54345ce8169e21184">get_cl_type_from_data_type</a>(dt));</div><div class="line"><a name="l00241"></a><span class="lineno"> 241</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(dt == <a class="code" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58a56d8353718e6fdc78b8d69078a2cdb94">DataType::F16</a>, <span class="stringliteral">"-DUSE_F16"</span>);</div><div class="line"><a name="l00242"></a><span class="lineno"> 242</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(<a class="code" href="namespacearm__compute.xhtml#af5982a092e9eb743fce2d6392bdd8897">is_data_type_float</a>(dt) && (beta != 1.0f), <span class="stringliteral">"-DBETA="</span> + <a class="code" href="namespacearm__compute.xhtml#a9cd394c15b73f79ca1d98f5328064be2">float_to_string_with_full_precision</a>(beta));</div><div class="line"><a name="l00243"></a><span class="lineno"> 243</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a504e7a51a86c19f1f529f739f50729f2">add_options_if</a>(<a class="code" href="namespacearm__compute.xhtml#a14f46283f316e7f0fad301d5c1507e9f">is_data_type_quantized_asymmetric</a>(dt), prepare_quantized_softmax_build_options(input-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a3f3e1a3200223e6a304a533b1016e749">quantization_info</a>().<a class="code" href="structarm__compute_1_1_quantization_info.xhtml#a1d28dec57cce925ad92342891bd71e7c">scale</a>, beta).options());</div><div class="line"><a name="l00244"></a><span class="lineno"> 244</span> </div><div class="line"><a name="l00245"></a><span class="lineno"> 245</span>  cl::NDRange <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ab9f813c25ed75ea7b7ac2fa3926a8f55">lws_hint</a>(cl::NullRange);</div><div class="line"><a name="l00246"></a><span class="lineno"> 246</span>  std::string kernel_name = <a class="code" href="namespacearm__compute.xhtml#a14f46283f316e7f0fad301d5c1507e9f">is_data_type_quantized_asymmetric</a>(dt) ? std::string(<span class="stringliteral">"softmax_layer_max_shift_exp_sum_quantized_serial"</span>) :</div><div class="line"><a name="l00247"></a><span class="lineno"> 247</span>  std::string(<span class="stringliteral">"softmax_layer_max_shift_exp_sum_serial"</span>);</div><div class="line"><a name="l00248"></a><span class="lineno"> 248</span>  <a class="code" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a6474b575b0139e3e1973acc9179556b8">ParallelReductionInfo</a> parallel_reduction_info = <a class="code" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a78a33824fc1fd7c44535530ebd917eb4">is_parallel_reduction</a>(reduction_dim_size);</div><div class="line"><a name="l00249"></a><span class="lineno"> 249</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> vector_size = std::get<1>(parallel_reduction_info);</div><div class="line"><a name="l00250"></a><span class="lineno"> 250</span> </div><div class="line"><a name="l00251"></a><span class="lineno"> 251</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a3e2b80ff5463b7d2017de847f5c32a30">add_option</a>(<span class="stringliteral">"-DVECTOR_SIZE="</span> + <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(vector_size));</div><div class="line"><a name="l00252"></a><span class="lineno"> 252</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a3e2b80ff5463b7d2017de847f5c32a30">add_option</a>(<span class="stringliteral">"-DLOG_VECTOR_SIZE="</span> + <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(lround(log2(vector_size))));</div><div class="line"><a name="l00253"></a><span class="lineno"> 253</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>((reduction_dim_size % vector_size) != 0, <span class="stringliteral">"-DNON_MULTIPLE_OF_VECTOR_SIZE"</span>);</div><div class="line"><a name="l00254"></a><span class="lineno"> 254</span> </div><div class="line"><a name="l00255"></a><span class="lineno"> 255</span>  <span class="comment">// Configure parallel kernel if needed</span></div><div class="line"><a name="l00256"></a><span class="lineno"> 256</span>  <span class="keywordflow">if</span>(std::get<0>(parallel_reduction_info))</div><div class="line"><a name="l00257"></a><span class="lineno"> 257</span>  {</div><div class="line"><a name="l00258"></a><span class="lineno"> 258</span>  kernel_name = <a class="code" href="namespacearm__compute.xhtml#a14f46283f316e7f0fad301d5c1507e9f">is_data_type_quantized_asymmetric</a>(dt) ? std::string(<span class="stringliteral">"softmax_layer_max_shift_exp_sum_quantized_parallel"</span>) : std::string(<span class="stringliteral">"softmax_layer_max_shift_exp_sum_parallel"</span>);</div><div class="line"><a name="l00259"></a><span class="lineno"> 259</span>  <span class="keywordtype">bool</span> is_grid_size_pow2 = (_grid_size != 0) && ((_grid_size & (_grid_size - 1)) == 0);</div><div class="line"><a name="l00260"></a><span class="lineno"> 260</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>(is_grid_size_pow2 && _grid_size <= 256, <span class="stringliteral">"-DGRID_SIZE="</span> + <a class="code" href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">support::cpp11::to_string</a>(_grid_size));</div><div class="line"><a name="l00261"></a><span class="lineno"> 261</span> </div><div class="line"><a name="l00262"></a><span class="lineno"> 262</span>  <span class="comment">// Handle boundary conditions.</span></div><div class="line"><a name="l00263"></a><span class="lineno"> 263</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> multiple_grid_size = (reduction_dim_size / vector_size) % _grid_size;</div><div class="line"><a name="l00264"></a><span class="lineno"> 264</span>  build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">add_option_if</a>((multiple_grid_size != 0) || ((reduction_dim_size % vector_size) != 0), <span class="stringliteral">"-DNON_MULTIPLE_OF_GRID_SIZE"</span>);</div><div class="line"><a name="l00265"></a><span class="lineno"> 265</span>  <span class="comment">// Setting _lws_hint in this way can also communicate grid_size to CLLogits1DMaxShiftExpSumKernel::run().</span></div><div class="line"><a name="l00266"></a><span class="lineno"> 266</span>  <span class="comment">// A single workgroup performs reduction in dimension 0 in the parallel case, hence lws[0]==gws[0].</span></div><div class="line"><a name="l00267"></a><span class="lineno"> 267</span>  <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ab9f813c25ed75ea7b7ac2fa3926a8f55">lws_hint</a> = cl::NDRange(_grid_size);</div><div class="line"><a name="l00268"></a><span class="lineno"> 268</span>  }</div><div class="line"><a name="l00269"></a><span class="lineno"> 269</span> </div><div class="line"><a name="l00270"></a><span class="lineno"> 270</span>  <span class="comment">// Create kernel.</span></div><div class="line"><a name="l00271"></a><span class="lineno"> 271</span>  _kernel = static_cast<cl::Kernel>(<a class="code" href="classarm__compute_1_1_c_l_kernel_library.xhtml#acba005f5ce2c62cbf3f94d074d9007aa">CLKernelLibrary::get</a>().<a class="code" href="namespacearm__compute.xhtml#abc72c95941485d8a068fa38372308574">create_kernel</a>(kernel_name, build_opts.<a class="code" href="classarm__compute_1_1_c_l_build_options.xhtml#ae3b08139a1e57323c5d7dd93f30496c8">options</a>()));</div><div class="line"><a name="l00272"></a><span class="lineno"> 272</span> </div><div class="line"><a name="l00273"></a><span class="lineno"> 273</span>  <span class="comment">// Set static arguments. Both the kernels use the same arguments</span></div><div class="line"><a name="l00274"></a><span class="lineno"> 274</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> idx = 4 * <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a6c9c1e7a7d96743375ca40847f0f12e2">num_arguments_per_3D_tensor</a>(); <span class="comment">//Skip the input and output parameters</span></div><div class="line"><a name="l00275"></a><span class="lineno"> 275</span>  _kernel.setArg<cl_uint>(idx++, reduction_dim_size);</div><div class="line"><a name="l00276"></a><span class="lineno"> 276</span> </div><div class="line"><a name="l00277"></a><span class="lineno"> 277</span>  <span class="comment">// Configure window</span></div><div class="line"><a name="l00278"></a><span class="lineno"> 278</span>  <span class="keyword">auto</span> win_config = validate_and_configure_window_1DMaxShiftExpSum(input-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>(), max-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>(), output-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>(), <a class="code" href="reduction__operation_8cl.xhtml#ab0df00f5333da51860deb93deb44a782">sum</a>->info());</div><div class="line"><a name="l00279"></a><span class="lineno"> 279</span>  <a class="code" href="_error_8h.xhtml#a938dcd406ce611ef5345ad2531cdb948">ARM_COMPUTE_ERROR_THROW_ON</a>(win_config.first);</div><div class="line"><a name="l00280"></a><span class="lineno"> 280</span>  ICLKernel::configure_internal(win_config.second, <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ab9f813c25ed75ea7b7ac2fa3926a8f55">lws_hint</a>);</div><div class="line"><a name="l00281"></a><span class="lineno"> 281</span> }</div><div class="ttc" id="classarm__compute_1_1_i_tensor_info_xhtml_a178f0d3d87f959e00a743328d95359d2"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">arm_compute::ITensorInfo::dimension</a></div><div class="ttdeci">virtual size_t dimension(size_t index) const =0</div><div class="ttdoc">Return the size of the requested dimension.</div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_build_options_xhtml_ae3b08139a1e57323c5d7dd93f30496c8"><div class="ttname"><a href="classarm__compute_1_1_c_l_build_options.xhtml#ae3b08139a1e57323c5d7dd93f30496c8">arm_compute::CLBuildOptions::options</a></div><div class="ttdeci">const StringSet & options() const</div><div class="ttdoc">Gets the current options list set.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_kernel_library_8cpp_source.xhtml#l00074">CLKernelLibrary.cpp:74</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_c_l_kernel_xhtml_ab9f813c25ed75ea7b7ac2fa3926a8f55"><div class="ttname"><a href="classarm__compute_1_1_i_c_l_kernel.xhtml#ab9f813c25ed75ea7b7ac2fa3926a8f55">arm_compute::ICLKernel::lws_hint</a></div><div class="ttdeci">cl::NDRange lws_hint() const</div><div class="ttdoc">Return the Local-Workgroup-Size hint.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_kernel_8h_source.xhtml#l00219">ICLKernel.h:219</a></div></div> |
| <div class="ttc" id="reduction__operation_8cl_xhtml_ab0df00f5333da51860deb93deb44a782"><div class="ttname"><a href="reduction__operation_8cl.xhtml#ab0df00f5333da51860deb93deb44a782">sum</a></div><div class="ttdeci">DATA_TYPE sum(__global const DATA_TYPE *input)</div><div class="ttdoc">Calculate sum of a vector.</div><div class="ttdef"><b>Definition:</b> <a href="reduction__operation_8cl_source.xhtml#l00052">reduction_operation.cl:52</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1support_1_1cpp11_xhtml_acc5dddee1cbe93a4eaf0a9f74ee96bb7"><div class="ttname"><a href="namespacearm__compute_1_1support_1_1cpp11.xhtml#acc5dddee1cbe93a4eaf0a9f74ee96bb7">arm_compute::support::cpp11::to_string</a></div><div class="ttdeci">std::string to_string(T &&value)</div><div class="ttdoc">Convert integer and float values to string.</div><div class="ttdef"><b>Definition:</b> <a href="_toolchain_support_8h_source.xhtml#l00210">ToolchainSupport.h:210</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_info_xhtml_a7cfb31af63202568efef5214acfbf3ba"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_info.xhtml#a7cfb31af63202568efef5214acfbf3ba">arm_compute::ITensorInfo::data_type</a></div><div class="ttdeci">virtual DataType data_type() const =0</div><div class="ttdoc">Data type used for each element of the tensor.</div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_kernel_library_xhtml_acba005f5ce2c62cbf3f94d074d9007aa"><div class="ttname"><a href="classarm__compute_1_1_c_l_kernel_library.xhtml#acba005f5ce2c62cbf3f94d074d9007aa">arm_compute::CLKernelLibrary::get</a></div><div class="ttdeci">static CLKernelLibrary & get()</div><div class="ttdoc">Access the KernelLibrary singleton.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_kernel_library_8cpp_source.xhtml#l00958">CLKernelLibrary.cpp:958</a></div></div> |
| <div class="ttc" id="_error_8h_xhtml_a938dcd406ce611ef5345ad2531cdb948"><div class="ttname"><a href="_error_8h.xhtml#a938dcd406ce611ef5345ad2531cdb948">ARM_COMPUTE_ERROR_THROW_ON</a></div><div class="ttdeci">#define ARM_COMPUTE_ERROR_THROW_ON(status)</div><div class="ttdef"><b>Definition:</b> <a href="_error_8h_source.xhtml#l00327">Error.h:327</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a47be6fa38308d0003c25b60b7dbc45ce"><div class="ttname"><a href="namespacearm__compute.xhtml#a47be6fa38308d0003c25b60b7dbc45ce">arm_compute::auto_init_if_empty</a></div><div class="ttdeci">bool auto_init_if_empty(ITensorInfo &info, const TensorShape &shape, int num_channels, DataType data_type, QuantizationInfo quantization_info=QuantizationInfo())</div><div class="ttdoc">Auto initialize the tensor info (shape, number of channels and data type) if the current assignment i...</div><div class="ttdef"><b>Definition:</b> <a href="_helpers_8inl_source.xhtml#l00201">Helpers.inl:201</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_ab4e88c89b3b7ea1735996cc4def22d58a56d8353718e6fdc78b8d69078a2cdb94"><div class="ttname"><a href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58a56d8353718e6fdc78b8d69078a2cdb94">arm_compute::Format::F16</a></div><div class="ttdoc">1 channel, 1 F16 per channel</div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_build_options_xhtml_a3e2b80ff5463b7d2017de847f5c32a30"><div class="ttname"><a href="classarm__compute_1_1_c_l_build_options.xhtml#a3e2b80ff5463b7d2017de847f5c32a30">arm_compute::CLBuildOptions::add_option</a></div><div class="ttdeci">void add_option(std::string option)</div><div class="ttdoc">Adds option to the existing build option list.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_kernel_library_8cpp_source.xhtml#l00043">CLKernelLibrary.cpp:43</a></div></div> |
| <div class="ttc" id="structarm__compute_1_1_quantization_info_xhtml_a1d28dec57cce925ad92342891bd71e7c"><div class="ttname"><a href="structarm__compute_1_1_quantization_info.xhtml#a1d28dec57cce925ad92342891bd71e7c">arm_compute::QuantizationInfo::scale</a></div><div class="ttdeci">float scale</div><div class="ttdoc">scale</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l00189">Types.h:189</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_c_l_kernel_xhtml_a6c9c1e7a7d96743375ca40847f0f12e2"><div class="ttname"><a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a6c9c1e7a7d96743375ca40847f0f12e2">arm_compute::ICLKernel::num_arguments_per_3D_tensor</a></div><div class="ttdeci">static constexpr unsigned int num_arguments_per_3D_tensor()</div><div class="ttdoc">Returns the number of arguments enqueued per 3D tensor object.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_kernel_8h_source.xhtml#l00172">ICLKernel.h:172</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a9cd394c15b73f79ca1d98f5328064be2"><div class="ttname"><a href="namespacearm__compute.xhtml#a9cd394c15b73f79ca1d98f5328064be2">arm_compute::float_to_string_with_full_precision</a></div><div class="ttdeci">std::string float_to_string_with_full_precision(float val)</div><div class="ttdoc">Create a string with the float in full precision.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_utils_8h_source.xhtml#l01031">Utils.h:1031</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_info_xhtml_a7c66505457d00ece3aa4b34cab80757d"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_info.xhtml#a7c66505457d00ece3aa4b34cab80757d">arm_compute::ITensorInfo::tensor_shape</a></div><div class="ttdeci">virtual const TensorShape & tensor_shape() const =0</div><div class="ttdoc">Size for each dimension of the tensor.</div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a545eeda2eaa3f5a54345ce8169e21184"><div class="ttname"><a href="namespacearm__compute.xhtml#a545eeda2eaa3f5a54345ce8169e21184">arm_compute::get_cl_type_from_data_type</a></div><div class="ttdeci">std::string get_cl_type_from_data_type(const DataType &dt)</div><div class="ttdoc">Translates a tensor data type to the appropriate OpenCL type.</div><div class="ttdef"><b>Definition:</b> <a href="core_2_c_l_2_c_l_helpers_8cpp_source.xhtml#l00035">CLHelpers.cpp:35</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1misc_1_1_i_cloneable_xhtml_a4d10e5012a872e7f78f2b539b673049d"><div class="ttname"><a href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">arm_compute::misc::ICloneable::clone</a></div><div class="ttdeci">virtual std::unique_ptr< T > clone() const =0</div><div class="ttdoc">Provide a clone of the current object of class T.</div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_xhtml_a0e95dc1e53c361348314873b168ae237"><div class="ttname"><a href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">arm_compute::ITensor::info</a></div><div class="ttdeci">virtual ITensorInfo * info() const =0</div><div class="ttdoc">Interface to be implemented by the child class to return the tensor's metadata.</div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_build_options_xhtml_a95b46e69297fad10b27a1baa000f92cc"><div class="ttname"><a href="classarm__compute_1_1_c_l_build_options.xhtml#a95b46e69297fad10b27a1baa000f92cc">arm_compute::CLBuildOptions::add_option_if</a></div><div class="ttdeci">void add_option_if(bool cond, std::string option)</div><div class="ttdoc">Adds option if a given condition is true;.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_kernel_library_8cpp_source.xhtml#l00048">CLKernelLibrary.cpp:48</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_build_options_xhtml"><div class="ttname"><a href="classarm__compute_1_1_c_l_build_options.xhtml">arm_compute::CLBuildOptions</a></div><div class="ttdoc">Build options.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_kernel_library_8h_source.xhtml#l00037">CLKernelLibrary.h:37</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_abc72c95941485d8a068fa38372308574"><div class="ttname"><a href="namespacearm__compute.xhtml#abc72c95941485d8a068fa38372308574">arm_compute::create_kernel</a></div><div class="ttdeci">std::unique_ptr< Kernel > create_kernel()</div><div class="ttdoc">Helper function to create and return a unique_ptr pointed to a CL/GLES kernel object.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_helpers_8h_source.xhtml#l00086">Helpers.h:86</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_info_xhtml_a3f3e1a3200223e6a304a533b1016e749"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_info.xhtml#a3f3e1a3200223e6a304a533b1016e749">arm_compute::ITensorInfo::quantization_info</a></div><div class="ttdeci">virtual QuantizationInfo quantization_info() const =0</div><div class="ttdoc">Get the quantization settings (scale and offset) of the tensor.</div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a14f46283f316e7f0fad301d5c1507e9f"><div class="ttname"><a href="namespacearm__compute.xhtml#a14f46283f316e7f0fad301d5c1507e9f">arm_compute::is_data_type_quantized_asymmetric</a></div><div class="ttdeci">bool is_data_type_quantized_asymmetric(DataType dt)</div><div class="ttdoc">Check if a given data type is of asymmetric quantized type.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_utils_8h_source.xhtml#l01014">Utils.h:1014</a></div></div> |
| <div class="ttc" id="_validate_8h_xhtml_a921b705e9e3e0fe928928447869e62a5"><div class="ttname"><a href="_validate_8h.xhtml#a921b705e9e3e0fe928928447869e62a5">ARM_COMPUTE_ERROR_ON_NULLPTR</a></div><div class="ttdeci">#define ARM_COMPUTE_ERROR_ON_NULLPTR(...)</div><div class="ttdef"><b>Definition:</b> <a href="_validate_8h_source.xhtml#l00161">Validate.h:161</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel_xhtml_a78a33824fc1fd7c44535530ebd917eb4"><div class="ttname"><a href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a78a33824fc1fd7c44535530ebd917eb4">arm_compute::CLLogits1DMaxShiftExpSumKernel::is_parallel_reduction</a></div><div class="ttdeci">static ParallelReductionInfo is_parallel_reduction(size_t size)</div><div class="ttdoc">Checks if the given size is eligible for parallel reduction.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_softmax_layer_kernel_8cpp_source.xhtml#l00291">CLSoftmaxLayerKernel.cpp:291</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel_xhtml_a6474b575b0139e3e1973acc9179556b8"><div class="ttname"><a href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a6474b575b0139e3e1973acc9179556b8">arm_compute::CLLogits1DMaxShiftExpSumKernel::ParallelReductionInfo</a></div><div class="ttdeci">std::tuple< bool, unsigned int > ParallelReductionInfo</div><div class="ttdoc">Info for whether a parallel reduction will be run and the vector size of the execution.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_softmax_layer_kernel_8h_source.xhtml#l00104">CLSoftmaxLayerKernel.h:104</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_ad8ed01ff3ff33333d8e19db4d2818bb6"><div class="ttname"><a href="namespacearm__compute.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6">arm_compute::DataType</a></div><div class="ttdeci">DataType</div><div class="ttdoc">Available data types.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_types_8h_source.xhtml#l00074">Types.h:74</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_af5982a092e9eb743fce2d6392bdd8897"><div class="ttname"><a href="namespacearm__compute.xhtml#af5982a092e9eb743fce2d6392bdd8897">arm_compute::is_data_type_float</a></div><div class="ttdeci">bool is_data_type_float(DataType dt)</div><div class="ttdoc">Check if a given data type is of floating point type.</div><div class="ttdef"><b>Definition:</b> <a href="arm__compute_2core_2_utils_8h_source.xhtml#l00977">Utils.h:977</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_build_options_xhtml_a504e7a51a86c19f1f529f739f50729f2"><div class="ttname"><a href="classarm__compute_1_1_c_l_build_options.xhtml#a504e7a51a86c19f1f529f739f50729f2">arm_compute::CLBuildOptions::add_options_if</a></div><div class="ttdeci">void add_options_if(bool cond, const StringSet &options)</div><div class="ttdoc">Appends given build options to the current's objects options if a given condition is true.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_kernel_library_8cpp_source.xhtml#l00066">CLKernelLibrary.cpp:66</a></div></div> |
| </div><!-- fragment --> |
| <p class="reference">References <a class="el" href="_c_l_kernel_library_8cpp_source.xhtml#l00043">CLBuildOptions::add_option()</a>, <a class="el" href="_c_l_kernel_library_8cpp_source.xhtml#l00048">CLBuildOptions::add_option_if()</a>, <a class="el" href="_c_l_kernel_library_8cpp_source.xhtml#l00066">CLBuildOptions::add_options_if()</a>, <a class="el" href="_validate_8h_source.xhtml#l00161">ARM_COMPUTE_ERROR_ON_NULLPTR</a>, <a class="el" href="_error_8h_source.xhtml#l00327">ARM_COMPUTE_ERROR_THROW_ON</a>, <a class="el" href="_helpers_8inl_source.xhtml#l00201">arm_compute::auto_init_if_empty()</a>, <a class="el" href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">ICloneable< T >::clone()</a>, <a class="el" href="classarm__compute_1_1_i_tensor_info.xhtml#a7cfb31af63202568efef5214acfbf3ba">ITensorInfo::data_type()</a>, <a class="el" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">ITensorInfo::dimension()</a>, <a class="el" href="namespacearm__compute.xhtml#ab4e88c89b3b7ea1735996cc4def22d58a56d8353718e6fdc78b8d69078a2cdb94">arm_compute::F16</a>, <a class="el" href="arm__compute_2core_2_utils_8h_source.xhtml#l01031">arm_compute::float_to_string_with_full_precision()</a>, <a class="el" href="core_2_c_l_2_c_l_helpers_8cpp_source.xhtml#l00035">arm_compute::get_cl_type_from_data_type()</a>, <a class="el" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">ITensor::info()</a>, <a class="el" href="arm__compute_2core_2_utils_8h_source.xhtml#l00977">arm_compute::is_data_type_float()</a>, <a class="el" href="arm__compute_2core_2_utils_8h_source.xhtml#l01014">arm_compute::is_data_type_quantized_asymmetric()</a>, <a class="el" href="_i_c_l_kernel_8h_source.xhtml#l00219">ICLKernel::lws_hint()</a>, <a class="el" href="classarm__compute_1_1_i_tensor_info.xhtml#a3f3e1a3200223e6a304a533b1016e749">ITensorInfo::quantization_info()</a>, <a class="el" href="arm__compute_2core_2_types_8h_source.xhtml#l00189">QuantizationInfo::scale</a>, <a class="el" href="reduction__operation_8cl_source.xhtml#l00052">sum()</a>, and <a class="el" href="classarm__compute_1_1_i_tensor_info.xhtml#a7c66505457d00ece3aa4b34cab80757d">ITensorInfo::tensor_shape()</a>.</p> |
| |
| <p class="reference">Referenced by <a class="el" href="_c_l_softmax_layer_8cpp_source.xhtml#l00073">CLSoftmaxLayer::configure()</a>.</p> |
| |
| </div> |
| </div> |
| <a id="a78a33824fc1fd7c44535530ebd917eb4"></a> |
| <h2 class="memtitle"><span class="permalink"><a href="#a78a33824fc1fd7c44535530ebd917eb4">◆ </a></span>is_parallel_reduction()</h2> |
| |
| <div class="memitem"> |
| <div class="memproto"> |
| <table class="mlabels"> |
| <tr> |
| <td class="mlabels-left"> |
| <table class="memname"> |
| <tr> |
| <td class="memname"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a6474b575b0139e3e1973acc9179556b8">CLLogits1DMaxShiftExpSumKernel::ParallelReductionInfo</a> is_parallel_reduction </td> |
| <td>(</td> |
| <td class="paramtype">size_t </td> |
| <td class="paramname"><em>size</em></td><td>)</td> |
| <td></td> |
| </tr> |
| </table> |
| </td> |
| <td class="mlabels-right"> |
| <span class="mlabels"><span class="mlabel">static</span></span> </td> |
| </tr> |
| </table> |
| </div><div class="memdoc"> |
| |
| <p>Checks if the given size is eligible for parallel reduction. </p> |
| <dl class="section note"><dt>Note</dt><dd>Serial reduction is launched for width < (_grid_size * _serial_vector_size). </dd> |
| <dd> |
| Parallel reduction is launched for width >= (_grid_size * _serial_vector_size) and vector_size is forced to 4.</dd></dl> |
| <dl class="params"><dt>Parameters</dt><dd> |
| <table class="params"> |
| <tr><td class="paramdir">[in]</td><td class="paramname">size</td><td>Size to check</td></tr> |
| </table> |
| </dd> |
| </dl> |
| <dl class="section return"><dt>Returns</dt><dd>A two-element tuple where the first element is a boolean specifying if a parallel reduction will be run, while the second element is the vector size of the execution. </dd></dl> |
| |
| <p class="definition">Definition at line <a class="el" href="_c_l_softmax_layer_kernel_8cpp_source.xhtml#l00291">291</a> of file <a class="el" href="_c_l_softmax_layer_kernel_8cpp_source.xhtml">CLSoftmaxLayerKernel.cpp</a>.</p> |
| <div class="fragment"><div class="line"><a name="l00292"></a><span class="lineno"> 292</span> {</div><div class="line"><a name="l00293"></a><span class="lineno"> 293</span>  <span class="keywordtype">bool</span> <a class="code" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a78a33824fc1fd7c44535530ebd917eb4">is_parallel_reduction</a> = (size >= (_grid_size * _serial_vector_size)) && (_grid_size > 1);</div><div class="line"><a name="l00294"></a><span class="lineno"> 294</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> vector_size = <a class="code" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a78a33824fc1fd7c44535530ebd917eb4">is_parallel_reduction</a> ? _parallel_vector_size : _serial_vector_size;</div><div class="line"><a name="l00295"></a><span class="lineno"> 295</span>  <span class="keywordflow">return</span> std::make_tuple(<a class="code" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a78a33824fc1fd7c44535530ebd917eb4">is_parallel_reduction</a>, vector_size);</div><div class="line"><a name="l00296"></a><span class="lineno"> 296</span> }</div><div class="ttc" id="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel_xhtml_a78a33824fc1fd7c44535530ebd917eb4"><div class="ttname"><a href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a78a33824fc1fd7c44535530ebd917eb4">arm_compute::CLLogits1DMaxShiftExpSumKernel::is_parallel_reduction</a></div><div class="ttdeci">static ParallelReductionInfo is_parallel_reduction(size_t size)</div><div class="ttdoc">Checks if the given size is eligible for parallel reduction.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_softmax_layer_kernel_8cpp_source.xhtml#l00291">CLSoftmaxLayerKernel.cpp:291</a></div></div> |
| </div><!-- fragment --> |
| <p class="reference">Referenced by <a class="el" href="validation_2_c_l_2_softmax_layer_8cpp_source.xhtml#l00065">arm_compute::test::validation::DATA_TEST_CASE()</a>, and <a class="el" href="_c_l_softmax_layer_kernel_8cpp_source.xhtml#l00298">CLLogits1DMaxShiftExpSumKernel::run()</a>.</p> |
| |
| </div> |
| </div> |
| <a id="afd28bfde2ba6f90c161eb1f2678cb6f9"></a> |
| <h2 class="memtitle"><span class="permalink"><a href="#afd28bfde2ba6f90c161eb1f2678cb6f9">◆ </a></span>operator=() <span class="overload">[1/2]</span></h2> |
| |
| <div class="memitem"> |
| <div class="memproto"> |
| <table class="mlabels"> |
| <tr> |
| <td class="mlabels-left"> |
| <table class="memname"> |
| <tr> |
| <td class="memname"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a>& operator= </td> |
| <td>(</td> |
| <td class="paramtype">const <a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a> & </td> |
| <td class="paramname"></td><td>)</td> |
| <td></td> |
| </tr> |
| </table> |
| </td> |
| <td class="mlabels-right"> |
| <span class="mlabels"><span class="mlabel">delete</span></span> </td> |
| </tr> |
| </table> |
| </div><div class="memdoc"> |
| |
| <p>Prevent instances of this class from being copied (As this class contains pointers) </p> |
| |
| </div> |
| </div> |
| <a id="ac7ea34435f64f3e3dce08ab0ce67295e"></a> |
| <h2 class="memtitle"><span class="permalink"><a href="#ac7ea34435f64f3e3dce08ab0ce67295e">◆ </a></span>operator=() <span class="overload">[2/2]</span></h2> |
| |
| <div class="memitem"> |
| <div class="memproto"> |
| <table class="mlabels"> |
| <tr> |
| <td class="mlabels-left"> |
| <table class="memname"> |
| <tr> |
| <td class="memname"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a>& operator= </td> |
| <td>(</td> |
| <td class="paramtype"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a> && </td> |
| <td class="paramname"></td><td>)</td> |
| <td></td> |
| </tr> |
| </table> |
| </td> |
| <td class="mlabels-right"> |
| <span class="mlabels"><span class="mlabel">default</span></span> </td> |
| </tr> |
| </table> |
| </div><div class="memdoc"> |
| |
| <p>Allow instances of this class to be moved. </p> |
| |
| </div> |
| </div> |
| <a id="a493987e85723a8000eb26d1f00e2ad0e"></a> |
| <h2 class="memtitle"><span class="permalink"><a href="#a493987e85723a8000eb26d1f00e2ad0e">◆ </a></span>run()</h2> |
| |
| <div class="memitem"> |
| <div class="memproto"> |
| <table class="mlabels"> |
| <tr> |
| <td class="mlabels-left"> |
| <table class="memname"> |
| <tr> |
| <td class="memname">void run </td> |
| <td>(</td> |
| <td class="paramtype">const <a class="el" href="classarm__compute_1_1_window.xhtml">Window</a> & </td> |
| <td class="paramname"><em>window</em>, </td> |
| </tr> |
| <tr> |
| <td class="paramkey"></td> |
| <td></td> |
| <td class="paramtype">cl::CommandQueue & </td> |
| <td class="paramname"><em>queue</em> </td> |
| </tr> |
| <tr> |
| <td></td> |
| <td>)</td> |
| <td></td><td></td> |
| </tr> |
| </table> |
| </td> |
| <td class="mlabels-right"> |
| <span class="mlabels"><span class="mlabel">override</span><span class="mlabel">virtual</span></span> </td> |
| </tr> |
| </table> |
| </div><div class="memdoc"> |
| |
| <p>Enqueue the OpenCL kernel to process the given window on the passed OpenCL command queue. </p> |
| <dl class="section note"><dt>Note</dt><dd>The queue is <em>not</em> flushed by this method, and therefore the kernel will not have been executed by the time this method returns.</dd></dl> |
| <dl class="params"><dt>Parameters</dt><dd> |
| <table class="params"> |
| <tr><td class="paramdir">[in]</td><td class="paramname">window</td><td>Region on which to execute the kernel. (Must be a valid region of the window returned by <a class="el" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6" title="The maximum window the kernel can be executed on.">window()</a>). </td></tr> |
| <tr><td class="paramdir">[in,out]</td><td class="paramname">queue</td><td>Command queue on which to enqueue the kernel. </td></tr> |
| </table> |
| </dd> |
| </dl> |
| |
| <p>Implements <a class="el" href="classarm__compute_1_1_i_c_l_kernel.xhtml#af6a174d47571f51f199ffc27ecc10f51">ICLKernel</a>.</p> |
| |
| <p class="definition">Definition at line <a class="el" href="_c_l_softmax_layer_kernel_8cpp_source.xhtml#l00298">298</a> of file <a class="el" href="_c_l_softmax_layer_kernel_8cpp_source.xhtml">CLSoftmaxLayerKernel.cpp</a>.</p> |
| <div class="fragment"><div class="line"><a name="l00299"></a><span class="lineno"> 299</span> {</div><div class="line"><a name="l00300"></a><span class="lineno"> 300</span>  <a class="code" href="_validate_8h.xhtml#a1b35b0d258183cf9ef36adf684d0b88c">ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL</a>(<span class="keyword">this</span>);</div><div class="line"><a name="l00301"></a><span class="lineno"> 301</span>  <a class="code" href="_validate_8h.xhtml#a6eb9ce82815fe429250189da7592ba75">ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW</a>(<a class="code" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">IKernel::window</a>(), <a class="code" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a>);</div><div class="line"><a name="l00302"></a><span class="lineno"> 302</span> </div><div class="line"><a name="l00303"></a><span class="lineno"> 303</span>  <span class="comment">// Collapse window in Z dimension</span></div><div class="line"><a name="l00304"></a><span class="lineno"> 304</span>  <a class="code" href="classarm__compute_1_1_window.xhtml">Window</a> window_collapsed = <a class="code" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">window</a>.<a class="code" href="classarm__compute_1_1_window.xhtml#a663ce71e1992366180fcb85b0a874e2f">collapse_if_possible</a>(<a class="code" href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">ICLKernel::window</a>(), <a class="code" href="classarm__compute_1_1_window.xhtml#a893d17b56b9abc4423ce26e9a24ac5dc">Window::DimZ</a>);</div><div class="line"><a name="l00305"></a><span class="lineno"> 305</span> </div><div class="line"><a name="l00306"></a><span class="lineno"> 306</span>  <span class="comment">// Reconfigure window in case of parallel reduction</span></div><div class="line"><a name="l00307"></a><span class="lineno"> 307</span>  <a class="code" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a6474b575b0139e3e1973acc9179556b8">ParallelReductionInfo</a> parallel_reduction_info = <a class="code" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a78a33824fc1fd7c44535530ebd917eb4">is_parallel_reduction</a>(_input-><a class="code" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">info</a>()-><a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">dimension</a>(0));</div><div class="line"><a name="l00308"></a><span class="lineno"> 308</span>  <span class="keywordflow">if</span>(std::get<0>(parallel_reduction_info))</div><div class="line"><a name="l00309"></a><span class="lineno"> 309</span>  {</div><div class="line"><a name="l00310"></a><span class="lineno"> 310</span>  <span class="comment">// To launch grid_size parallel workitems, steps.x should be modified as follows.</span></div><div class="line"><a name="l00311"></a><span class="lineno"> 311</span>  <span class="keyword">const</span> <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a75b73e17c4ebe901e44af3b2b9846ab3">step</a> = std::get<1>(parallel_reduction_info);</div><div class="line"><a name="l00312"></a><span class="lineno"> 312</span>  window_collapsed.<a class="code" href="classarm__compute_1_1_window.xhtml#acd3d2bba51cb84d34dd7656ad2375a6e">set</a>(<a class="code" href="classarm__compute_1_1_window.xhtml#aa96e81276ee4f87ab386cd05a5539a7d">Window::DimX</a>, <a class="code" href="classarm__compute_1_1_window_1_1_dimension.xhtml">Window::Dimension</a>(0, _grid_size * <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a75b73e17c4ebe901e44af3b2b9846ab3">step</a>, <a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a75b73e17c4ebe901e44af3b2b9846ab3">step</a>));</div><div class="line"><a name="l00313"></a><span class="lineno"> 313</span>  }</div><div class="line"><a name="l00314"></a><span class="lineno"> 314</span> </div><div class="line"><a name="l00315"></a><span class="lineno"> 315</span>  <span class="comment">// Get slices</span></div><div class="line"><a name="l00316"></a><span class="lineno"> 316</span>  <a class="code" href="classarm__compute_1_1_window.xhtml">Window</a> <a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a> = window_collapsed.<a class="code" href="classarm__compute_1_1_window.xhtml#a30ca5bdbb60ee281d7f1ab34f7a4ee40">first_slice_window_3D</a>();</div><div class="line"><a name="l00317"></a><span class="lineno"> 317</span>  <span class="keywordflow">do</span></div><div class="line"><a name="l00318"></a><span class="lineno"> 318</span>  {</div><div class="line"><a name="l00319"></a><span class="lineno"> 319</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> idx = 0;</div><div class="line"><a name="l00320"></a><span class="lineno"> 320</span>  <span class="comment">// Set inputs</span></div><div class="line"><a name="l00321"></a><span class="lineno"> 321</span>  <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a28f5847162f352444c6ac1825d0e99c7">add_3D_tensor_argument</a>(idx, _input, <a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a>);</div><div class="line"><a name="l00322"></a><span class="lineno"> 322</span>  <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a28f5847162f352444c6ac1825d0e99c7">add_3D_tensor_argument</a>(idx, _max, <a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a>);</div><div class="line"><a name="l00323"></a><span class="lineno"> 323</span>  <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a28f5847162f352444c6ac1825d0e99c7">add_3D_tensor_argument</a>(idx, _output, <a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a>);</div><div class="line"><a name="l00324"></a><span class="lineno"> 324</span>  <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#a28f5847162f352444c6ac1825d0e99c7">add_3D_tensor_argument</a>(idx, _sum, <a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a>);</div><div class="line"><a name="l00325"></a><span class="lineno"> 325</span>  <a class="code" href="namespacearm__compute.xhtml#a1e48b7cf7de6a52575088cb7082d2b3e">enqueue</a>(queue, *<span class="keyword">this</span>, <a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a>, <a class="code" href="classarm__compute_1_1_i_c_l_kernel.xhtml#ab9f813c25ed75ea7b7ac2fa3926a8f55">lws_hint</a>());</div><div class="line"><a name="l00326"></a><span class="lineno"> 326</span>  }</div><div class="line"><a name="l00327"></a><span class="lineno"> 327</span>  <span class="keywordflow">while</span>(window_collapsed.<a class="code" href="classarm__compute_1_1_window.xhtml#aac792e3a11bc73bafafc4f4284c7f215">slide_window_slice_3D</a>(<a class="code" href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">slice</a>));</div><div class="line"><a name="l00328"></a><span class="lineno"> 328</span> }</div><div class="ttc" id="classarm__compute_1_1_i_kernel_xhtml_ad34a46f53686c12a5c5e717cc9617fb6"><div class="ttname"><a href="classarm__compute_1_1_i_kernel.xhtml#ad34a46f53686c12a5c5e717cc9617fb6">arm_compute::IKernel::window</a></div><div class="ttdeci">const Window & window() const</div><div class="ttdoc">The maximum window the kernel can be executed on.</div><div class="ttdef"><b>Definition:</b> <a href="_i_kernel_8cpp_source.xhtml#l00028">IKernel.cpp:28</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_info_xhtml_a178f0d3d87f959e00a743328d95359d2"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">arm_compute::ITensorInfo::dimension</a></div><div class="ttdeci">virtual size_t dimension(size_t index) const =0</div><div class="ttdoc">Return the size of the requested dimension.</div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_c_l_kernel_xhtml_ab9f813c25ed75ea7b7ac2fa3926a8f55"><div class="ttname"><a href="classarm__compute_1_1_i_c_l_kernel.xhtml#ab9f813c25ed75ea7b7ac2fa3926a8f55">arm_compute::ICLKernel::lws_hint</a></div><div class="ttdeci">cl::NDRange lws_hint() const</div><div class="ttdoc">Return the Local-Workgroup-Size hint.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_kernel_8h_source.xhtml#l00219">ICLKernel.h:219</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_window_1_1_dimension_xhtml"><div class="ttname"><a href="classarm__compute_1_1_window_1_1_dimension.xhtml">arm_compute::Window::Dimension</a></div><div class="ttdoc">Describe one of the image's dimensions with a start, end and step.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8h_source.xhtml#l00068">Window.h:68</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_c_l_kernel_xhtml_a28f5847162f352444c6ac1825d0e99c7"><div class="ttname"><a href="classarm__compute_1_1_i_c_l_kernel.xhtml#a28f5847162f352444c6ac1825d0e99c7">arm_compute::ICLKernel::add_3D_tensor_argument</a></div><div class="ttdeci">void add_3D_tensor_argument(unsigned int &idx, const ICLTensor *tensor, const Window &window)</div><div class="ttdoc">Add the passed 3D tensor's parameters to the object's kernel's arguments starting from the index idx.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_kernel_8h_source.xhtml#l00130">ICLKernel.h:130</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1test_1_1validation_xhtml_a75b73e17c4ebe901e44af3b2b9846ab3"><div class="ttname"><a href="namespacearm__compute_1_1test_1_1validation.xhtml#a75b73e17c4ebe901e44af3b2b9846ab3">arm_compute::test::validation::step</a></div><div class="ttdeci">const int step</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_2_select_8cpp_source.xhtml#l00172">Select.cpp:172</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_window_xhtml_aa96e81276ee4f87ab386cd05a5539a7d"><div class="ttname"><a href="classarm__compute_1_1_window.xhtml#aa96e81276ee4f87ab386cd05a5539a7d">arm_compute::Window::DimX</a></div><div class="ttdeci">static constexpr size_t DimX</div><div class="ttdoc">Alias for dimension 0 also known as X dimension.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8h_source.xhtml#l00043">Window.h:43</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_window_xhtml_a663ce71e1992366180fcb85b0a874e2f"><div class="ttname"><a href="classarm__compute_1_1_window.xhtml#a663ce71e1992366180fcb85b0a874e2f">arm_compute::Window::collapse_if_possible</a></div><div class="ttdeci">Window collapse_if_possible(const Window &full_window, size_t first, size_t last, bool *has_collapsed=nullptr) const</div><div class="ttdoc">Collapse the dimensions between first and last if possible.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8inl_source.xhtml#l00047">Window.inl:47</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_xhtml_a0e95dc1e53c361348314873b168ae237"><div class="ttname"><a href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">arm_compute::ITensor::info</a></div><div class="ttdeci">virtual ITensorInfo * info() const =0</div><div class="ttdoc">Interface to be implemented by the child class to return the tensor's metadata.</div></div> |
| <div class="ttc" id="classarm__compute_1_1_window_xhtml_acd3d2bba51cb84d34dd7656ad2375a6e"><div class="ttname"><a href="classarm__compute_1_1_window.xhtml#acd3d2bba51cb84d34dd7656ad2375a6e">arm_compute::Window::set</a></div><div class="ttdeci">void set(size_t dimension, const Dimension &dim)</div><div class="ttdoc">Set the values of a given dimension.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8inl_source.xhtml#l00041">Window.inl:41</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_window_xhtml_aac792e3a11bc73bafafc4f4284c7f215"><div class="ttname"><a href="classarm__compute_1_1_window.xhtml#aac792e3a11bc73bafafc4f4284c7f215">arm_compute::Window::slide_window_slice_3D</a></div><div class="ttdeci">bool slide_window_slice_3D(Window &slice) const</div><div class="ttdoc">Slide the passed 3D window slice.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8h_source.xhtml#l00312">Window.h:312</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml_a1e48b7cf7de6a52575088cb7082d2b3e"><div class="ttname"><a href="namespacearm__compute.xhtml#a1e48b7cf7de6a52575088cb7082d2b3e">arm_compute::enqueue</a></div><div class="ttdeci">void enqueue(cl::CommandQueue &queue, ICLKernel &kernel, const Window &window, const cl::NDRange &lws_hint=CLKernelLibrary::get().default_ndrange())</div><div class="ttdoc">Add the kernel to the command queue with the given window.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_kernel_8cpp_source.xhtml#l00039">ICLKernel.cpp:39</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel_xhtml_a78a33824fc1fd7c44535530ebd917eb4"><div class="ttname"><a href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a78a33824fc1fd7c44535530ebd917eb4">arm_compute::CLLogits1DMaxShiftExpSumKernel::is_parallel_reduction</a></div><div class="ttdeci">static ParallelReductionInfo is_parallel_reduction(size_t size)</div><div class="ttdoc">Checks if the given size is eligible for parallel reduction.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_softmax_layer_kernel_8cpp_source.xhtml#l00291">CLSoftmaxLayerKernel.cpp:291</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_window_xhtml_a893d17b56b9abc4423ce26e9a24ac5dc"><div class="ttname"><a href="classarm__compute_1_1_window.xhtml#a893d17b56b9abc4423ce26e9a24ac5dc">arm_compute::Window::DimZ</a></div><div class="ttdeci">static constexpr size_t DimZ</div><div class="ttdoc">Alias for dimension 2 also known as Z dimension.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8h_source.xhtml#l00047">Window.h:47</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel_xhtml_a6474b575b0139e3e1973acc9179556b8"><div class="ttname"><a href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml#a6474b575b0139e3e1973acc9179556b8">arm_compute::CLLogits1DMaxShiftExpSumKernel::ParallelReductionInfo</a></div><div class="ttdeci">std::tuple< bool, unsigned int > ParallelReductionInfo</div><div class="ttdoc">Info for whether a parallel reduction will be run and the vector size of the execution.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_softmax_layer_kernel_8h_source.xhtml#l00104">CLSoftmaxLayerKernel.h:104</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_window_xhtml_a30ca5bdbb60ee281d7f1ab34f7a4ee40"><div class="ttname"><a href="classarm__compute_1_1_window.xhtml#a30ca5bdbb60ee281d7f1ab34f7a4ee40">arm_compute::Window::first_slice_window_3D</a></div><div class="ttdeci">Window first_slice_window_3D() const</div><div class="ttdoc">First 3D slice of the window.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8h_source.xhtml#l00268">Window.h:268</a></div></div> |
| <div class="ttc" id="_validate_8h_xhtml_a6eb9ce82815fe429250189da7592ba75"><div class="ttname"><a href="_validate_8h.xhtml#a6eb9ce82815fe429250189da7592ba75">ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW</a></div><div class="ttdeci">#define ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW(f, s)</div><div class="ttdef"><b>Definition:</b> <a href="_validate_8h_source.xhtml#l00205">Validate.h:205</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_window_xhtml"><div class="ttname"><a href="classarm__compute_1_1_window.xhtml">arm_compute::Window</a></div><div class="ttdoc">Describe a multidimensional execution window.</div><div class="ttdef"><b>Definition:</b> <a href="_window_8h_source.xhtml#l00039">Window.h:39</a></div></div> |
| <div class="ttc" id="_validate_8h_xhtml_a1b35b0d258183cf9ef36adf684d0b88c"><div class="ttname"><a href="_validate_8h.xhtml#a1b35b0d258183cf9ef36adf684d0b88c">ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL</a></div><div class="ttdeci">#define ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL(k)</div><div class="ttdef"><b>Definition:</b> <a href="_validate_8h_source.xhtml#l00940">Validate.h:940</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1test_1_1validation_1_1reference_xhtml_a548131b3d37da47a2e9d32111c88dfe1"><div class="ttname"><a href="namespacearm__compute_1_1test_1_1validation_1_1reference.xhtml#a548131b3d37da47a2e9d32111c88dfe1">arm_compute::test::validation::reference::slice</a></div><div class="ttdeci">SimpleTensor< T > slice(const SimpleTensor< T > &src, Coordinates starts, Coordinates ends)</div><div class="ttdef"><b>Definition:</b> <a href="_slice_operations_8cpp_source.xhtml#l00038">SliceOperations.cpp:38</a></div></div> |
| </div><!-- fragment --> |
| <p class="reference">References <a class="el" href="_i_c_l_kernel_8h_source.xhtml#l00130">ICLKernel::add_3D_tensor_argument()</a>, <a class="el" href="_validate_8h_source.xhtml#l00205">ARM_COMPUTE_ERROR_ON_INVALID_SUBWINDOW</a>, <a class="el" href="_validate_8h_source.xhtml#l00940">ARM_COMPUTE_ERROR_ON_UNCONFIGURED_KERNEL</a>, <a class="el" href="_window_8inl_source.xhtml#l00047">Window::collapse_if_possible()</a>, <a class="el" href="classarm__compute_1_1_i_tensor_info.xhtml#a178f0d3d87f959e00a743328d95359d2">ITensorInfo::dimension()</a>, <a class="el" href="_window_8h_source.xhtml#l00043">Window::DimX</a>, <a class="el" href="_window_8h_source.xhtml#l00047">Window::DimZ</a>, <a class="el" href="_i_c_l_kernel_8cpp_source.xhtml#l00039">arm_compute::enqueue()</a>, <a class="el" href="_window_8h_source.xhtml#l00268">Window::first_slice_window_3D()</a>, <a class="el" href="classarm__compute_1_1_i_tensor.xhtml#a0e95dc1e53c361348314873b168ae237">ITensor::info()</a>, <a class="el" href="_c_l_softmax_layer_kernel_8cpp_source.xhtml#l00291">CLLogits1DMaxShiftExpSumKernel::is_parallel_reduction()</a>, <a class="el" href="_i_c_l_kernel_8h_source.xhtml#l00219">ICLKernel::lws_hint()</a>, <a class="el" href="_window_8inl_source.xhtml#l00041">Window::set()</a>, <a class="el" href="_slice_operations_8cpp_source.xhtml#l00038">arm_compute::test::validation::reference::slice()</a>, <a class="el" href="_window_8h_source.xhtml#l00312">Window::slide_window_slice_3D()</a>, <a class="el" href="_c_l_2_select_8cpp_source.xhtml#l00172">arm_compute::test::validation::step</a>, and <a class="el" href="_i_kernel_8cpp_source.xhtml#l00028">IKernel::window()</a>.</p> |
| |
| </div> |
| </div> |
| <a id="a62885f3bde4b458c02316bfc3f339fbc"></a> |
| <h2 class="memtitle"><span class="permalink"><a href="#a62885f3bde4b458c02316bfc3f339fbc">◆ </a></span>validate()</h2> |
| |
| <div class="memitem"> |
| <div class="memproto"> |
| <table class="mlabels"> |
| <tr> |
| <td class="mlabels-left"> |
| <table class="memname"> |
| <tr> |
| <td class="memname"><a class="el" href="classarm__compute_1_1_status.xhtml">Status</a> validate </td> |
| <td>(</td> |
| <td class="paramtype">const <a class="el" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> * </td> |
| <td class="paramname"><em>input</em>, </td> |
| </tr> |
| <tr> |
| <td class="paramkey"></td> |
| <td></td> |
| <td class="paramtype">const <a class="el" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> * </td> |
| <td class="paramname"><em>max</em>, </td> |
| </tr> |
| <tr> |
| <td class="paramkey"></td> |
| <td></td> |
| <td class="paramtype">const <a class="el" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> * </td> |
| <td class="paramname"><em>output</em>, </td> |
| </tr> |
| <tr> |
| <td class="paramkey"></td> |
| <td></td> |
| <td class="paramtype">const <a class="el" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> * </td> |
| <td class="paramname"><em>sum</em> </td> |
| </tr> |
| <tr> |
| <td></td> |
| <td>)</td> |
| <td></td><td></td> |
| </tr> |
| </table> |
| </td> |
| <td class="mlabels-right"> |
| <span class="mlabels"><span class="mlabel">static</span></span> </td> |
| </tr> |
| </table> |
| </div><div class="memdoc"> |
| |
| <p>Static function to check if given info will lead to a valid configuration of <a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a>. </p> |
| <dl class="params"><dt>Parameters</dt><dd> |
| <table class="params"> |
| <tr><td class="paramdir">[in]</td><td class="paramname">input</td><td>Source tensor. Data types supported: F16/F32 </td></tr> |
| <tr><td class="paramdir">[in]</td><td class="paramname">max</td><td>Max values tensor. Data types supported: same as <code>input</code> </td></tr> |
| <tr><td class="paramdir">[in]</td><td class="paramname">output</td><td>Destination tensor. Data types supported: same as <code>input</code> </td></tr> |
| <tr><td class="paramdir">[in]</td><td class="paramname">sum</td><td>Sum of 1D logits tensor. Data types supported: same as <code>input</code> </td></tr> |
| </table> |
| </dd> |
| </dl> |
| <dl class="section return"><dt>Returns</dt><dd>a status </dd></dl> |
| |
| <p class="definition">Definition at line <a class="el" href="_c_l_softmax_layer_kernel_8cpp_source.xhtml#l00283">283</a> of file <a class="el" href="_c_l_softmax_layer_kernel_8cpp_source.xhtml">CLSoftmaxLayerKernel.cpp</a>.</p> |
| <div class="fragment"><div class="line"><a name="l00284"></a><span class="lineno"> 284</span> {</div><div class="line"><a name="l00285"></a><span class="lineno"> 285</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(validate_arguments_1DMaxShiftExpSum(input, max, output, <a class="code" href="reduction__operation_8cl.xhtml#ab0df00f5333da51860deb93deb44a782">sum</a>));</div><div class="line"><a name="l00286"></a><span class="lineno"> 286</span>  <a class="code" href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a>(validate_and_configure_window_1DMaxShiftExpSum(input-><a class="code" href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">clone</a>().get(), max-><a class="code" href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">clone</a>().get(), output-><a class="code" href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">clone</a>().get(), <a class="code" href="reduction__operation_8cl.xhtml#ab0df00f5333da51860deb93deb44a782">sum</a>->clone().get()).first);</div><div class="line"><a name="l00287"></a><span class="lineno"> 287</span> </div><div class="line"><a name="l00288"></a><span class="lineno"> 288</span>  <span class="keywordflow">return</span> <a class="code" href="classarm__compute_1_1_status.xhtml">Status</a>{};</div><div class="line"><a name="l00289"></a><span class="lineno"> 289</span> }</div><div class="ttc" id="reduction__operation_8cl_xhtml_ab0df00f5333da51860deb93deb44a782"><div class="ttname"><a href="reduction__operation_8cl.xhtml#ab0df00f5333da51860deb93deb44a782">sum</a></div><div class="ttdeci">DATA_TYPE sum(__global const DATA_TYPE *input)</div><div class="ttdoc">Calculate sum of a vector.</div><div class="ttdef"><b>Definition:</b> <a href="reduction__operation_8cl_source.xhtml#l00052">reduction_operation.cl:52</a></div></div> |
| <div class="ttc" id="_error_8h_xhtml_a8a1e1c105f0bdaf37db408c7cfcb77a4"><div class="ttname"><a href="_error_8h.xhtml#a8a1e1c105f0bdaf37db408c7cfcb77a4">ARM_COMPUTE_RETURN_ON_ERROR</a></div><div class="ttdeci">#define ARM_COMPUTE_RETURN_ON_ERROR(status)</div><div class="ttdoc">Checks if a status contains an error and returns it.</div><div class="ttdef"><b>Definition:</b> <a href="_error_8h_source.xhtml#l00193">Error.h:193</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_status_xhtml"><div class="ttname"><a href="classarm__compute_1_1_status.xhtml">arm_compute::Status</a></div><div class="ttdoc">Status class.</div><div class="ttdef"><b>Definition:</b> <a href="_error_8h_source.xhtml#l00052">Error.h:52</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1misc_1_1_i_cloneable_xhtml_a4d10e5012a872e7f78f2b539b673049d"><div class="ttname"><a href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">arm_compute::misc::ICloneable::clone</a></div><div class="ttdeci">virtual std::unique_ptr< T > clone() const =0</div><div class="ttdoc">Provide a clone of the current object of class T.</div></div> |
| </div><!-- fragment --> |
| <p class="reference">References <a class="el" href="_error_8h_source.xhtml#l00193">ARM_COMPUTE_RETURN_ON_ERROR</a>, <a class="el" href="classarm__compute_1_1misc_1_1_i_cloneable.xhtml#a4d10e5012a872e7f78f2b539b673049d">ICloneable< T >::clone()</a>, and <a class="el" href="reduction__operation_8cl_source.xhtml#l00052">sum()</a>.</p> |
| |
| <p class="reference">Referenced by <a class="el" href="_c_l_softmax_layer_8cpp_source.xhtml#l00148">CLSoftmaxLayer::validate()</a>.</p> |
| |
| </div> |
| </div> |
| <hr/>The documentation for this class was generated from the following files:<ul> |
| <li>arm_compute/core/CL/kernels/<a class="el" href="_c_l_softmax_layer_kernel_8h_source.xhtml">CLSoftmaxLayerKernel.h</a></li> |
| <li>src/core/CL/kernels/<a class="el" href="_c_l_softmax_layer_kernel_8cpp_source.xhtml">CLSoftmaxLayerKernel.cpp</a></li> |
| </ul> |
| </div><!-- contents --> |
| </div><!-- doc-content --> |
| <!-- start footer part --> |
| <div id="nav-path" class="navpath"><!-- id is needed for treeview function! --> |
| <ul> |
| <li class="navelem"><a class="el" href="namespacearm__compute.xhtml">arm_compute</a></li><li class="navelem"><a class="el" href="classarm__compute_1_1_c_l_logits1_d_max_shift_exp_sum_kernel.xhtml">CLLogits1DMaxShiftExpSumKernel</a></li> |
| <li class="footer">Generated on Thu Feb 28 2019 12:25:00 for Compute Library by |
| <a href="http://www.doxygen.org/index.html"> |
| <img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.8.15 </li> |
| </ul> |
| </div> |
| </body> |
| </html> |