Documentation/_quantizer_test_8cpp_source.html

   1 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
   2 <html xmlns="http://www.w3.org/1999/xhtml">
   3 <head>
   4 <meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/>
   5 <meta http-equiv="X-UA-Compatible" content="IE=9"/>
   6 <meta name="generator" content="Doxygen 1.8.13"/>
   7 <meta name="viewport" content="width=device-width, initial-scale=1"/>
   8 <title>ArmNN: src/armnn/test/QuantizerTest.cpp Source File</title>
   9 <link href="tabs.css" rel="stylesheet" type="text/css"/>
  10 <script type="text/javascript" src="jquery.js"></script>
  11 <script type="text/javascript" src="dynsections.js"></script>
  12 <link href="navtree.css" rel="stylesheet" type="text/css"/>
  13 <script type="text/javascript" src="resize.js"></script>
  14 <script type="text/javascript" src="navtreedata.js"></script>
  15 <script type="text/javascript" src="navtree.js"></script>
  16 <script type="text/javascript">
  17   $(document).ready(initResizable);
  18 </script>
  19 <link href="search/search.css" rel="stylesheet" type="text/css"/>
  20 <script type="text/javascript" src="search/searchdata.js"></script>
  21 <script type="text/javascript" src="search/search.js"></script>
  22 <link href="doxygen.css" rel="stylesheet" type="text/css" />
  23 </head>
  24 <body>
  25 <div id="top"><!-- do not remove this div, it is closed by doxygen! -->
  26 <div id="titlearea">
  27 <table cellspacing="0" cellpadding="0">
  28  <tbody>
  29  <tr style="height: 56px;">
  30   <td id="projectalign" style="padding-left: 0.5em;">
  31    <div id="projectname">ArmNN
  32    &#160;<span id="projectnumber">NotReleased</span>
  33    </div>
  34   </td>
  35  </tr>
  36  </tbody>
  37 </table>
  38 </div>
  39 <!-- end header part -->
  40 <!-- Generated by Doxygen 1.8.13 -->
  41 <script type="text/javascript">
  42 var searchBox = new SearchBox("searchBox", "search",false,'Search');
  43 </script>
  44 <script type="text/javascript" src="menudata.js"></script>
  45 <script type="text/javascript" src="menu.js"></script>
  46 <script type="text/javascript">
  47 $(function() {
  48   initMenu('',true,false,'search.php','Search');
  49   $(document).ready(function() { init_search(); });
  50 });
  51 </script>
  52 <div id="main-nav"></div>
  53 </div><!-- top -->
  54 <div id="side-nav" class="ui-resizable side-nav-resizable">
  55   <div id="nav-tree">
  56     <div id="nav-tree-contents">
  57       <div id="nav-sync" class="sync"></div>
  58     </div>
  59   </div>
  60   <div id="splitbar" style="-moz-user-select:none;"
  61        class="ui-resizable-handle">
  62   </div>
  63 </div>
  64 <script type="text/javascript">
  65 $(document).ready(function(){initNavTree('_quantizer_test_8cpp_source.html','');});
  66 </script>
  67 <div id="doc-content">
  68 <!-- window showing the filter options -->
  69 <div id="MSearchSelectWindow"
  70      onmouseover="return searchBox.OnSearchSelectShow()"
  71      onmouseout="return searchBox.OnSearchSelectHide()"
  72      onkeydown="return searchBox.OnSearchSelectKey(event)">
  73 </div>
  74
  75 <!-- iframe showing the search results (closed by default) -->
  76 <div id="MSearchResultsWindow">
  77 <iframe src="javascript:void(0)" frameborder="0"
  78         name="MSearchResults" id="MSearchResults">
  79 </iframe>
  80 </div>
  81
  82 <div class="header">
  83   <div class="headertitle">
  84 <div class="title">QuantizerTest.cpp</div>  </div>
  85 </div><!--header-->
  86 <div class="contents">
  87 <a href="_quantizer_test_8cpp.html">Go to the documentation of this file.</a><div class="fragment"><div class="line"><a name="l00001"></a><span class="lineno">    1</span>&#160;<span class="comment">//</span></div><div class="line"><a name="l00002"></a><span class="lineno">    2</span>&#160;<span class="comment">// Copyright © 2017 Arm Ltd. All rights reserved.</span></div><div class="line"><a name="l00003"></a><span class="lineno">    3</span>&#160;<span class="comment">// SPDX-License-Identifier: MIT</span></div><div class="line"><a name="l00004"></a><span class="lineno">    4</span>&#160;<span class="comment">//</span></div><div class="line"><a name="l00005"></a><span class="lineno">    5</span>&#160;</div><div class="line"><a name="l00006"></a><span class="lineno">    6</span>&#160;<span class="preprocessor">#include &lt;<a class="code" href="_i_network_8hpp.html">armnn/INetwork.hpp</a>&gt;</span></div><div class="line"><a name="l00007"></a><span class="lineno">    7</span>&#160;<span class="preprocessor">#include &lt;<a class="code" href="_layer_visitor_base_8hpp.html">armnn/LayerVisitorBase.hpp</a>&gt;</span></div><div class="line"><a name="l00008"></a><span class="lineno">    8</span>&#160;<span class="preprocessor">#include &lt;<a class="code" href="_tensor_8hpp.html">armnn/Tensor.hpp</a>&gt;</span></div><div class="line"><a name="l00009"></a><span class="lineno">    9</span>&#160;<span class="preprocessor">#include &lt;<a class="code" href="_types_8hpp.html">armnn/Types.hpp</a>&gt;</span></div><div class="line"><a name="l00010"></a><span class="lineno">   10</span>&#160;</div><div class="line"><a name="l00011"></a><span class="lineno">   11</span>&#160;<span class="preprocessor">#include &lt;<a class="code" href="_i_network_quantizer_8hpp.html">armnnQuantizer/INetworkQuantizer.hpp</a>&gt;</span></div><div class="line"><a name="l00012"></a><span class="lineno">   12</span>&#160;</div><div class="line"><a name="l00013"></a><span class="lineno">   13</span>&#160;<span class="preprocessor">#include &lt;<a class="code" href="_quantize_helper_8hpp.html">QuantizeHelper.hpp</a>&gt;</span></div><div class="line"><a name="l00014"></a><span class="lineno">   14</span>&#160;</div><div class="line"><a name="l00015"></a><span class="lineno">   15</span>&#160;<span class="preprocessor">#include &quot;../Graph.hpp&quot;</span></div><div class="line"><a name="l00016"></a><span class="lineno">   16</span>&#160;<span class="preprocessor">#include &quot;../Network.hpp&quot;</span></div><div class="line"><a name="l00017"></a><span class="lineno">   17</span>&#160;<span class="preprocessor">#include &quot;../NetworkQuantizerUtils.hpp&quot;</span></div><div class="line"><a name="l00018"></a><span class="lineno">   18</span>&#160;<span class="preprocessor">#include &quot;../OverrideInputRangeVisitor.hpp&quot;</span></div><div class="line"><a name="l00019"></a><span class="lineno">   19</span>&#160;<span class="preprocessor">#include &quot;../RangeTracker.hpp&quot;</span></div><div class="line"><a name="l00020"></a><span class="lineno">   20</span>&#160;<span class="preprocessor">#include &quot;../../armnnQuantizer/CommandLineProcessor.hpp&quot;</span></div><div class="line"><a name="l00021"></a><span class="lineno">   21</span>&#160;</div><div class="line"><a name="l00022"></a><span class="lineno">   22</span>&#160;<span class="preprocessor">#include &lt;boost/core/ignore_unused.hpp&gt;</span></div><div class="line"><a name="l00023"></a><span class="lineno">   23</span>&#160;<span class="preprocessor">#include &lt;boost/test/unit_test.hpp&gt;</span></div><div class="line"><a name="l00024"></a><span class="lineno">   24</span>&#160;</div><div class="line"><a name="l00025"></a><span class="lineno">   25</span>&#160;<span class="preprocessor">#include &lt;unordered_map&gt;</span></div><div class="line"><a name="l00026"></a><span class="lineno">   26</span>&#160;</div><div class="line"><a name="l00027"></a><span class="lineno">   27</span>&#160;<span class="keyword">namespace </span><a class="code" href="namespacearmnn.html">armnn</a></div><div class="line"><a name="l00028"></a><span class="lineno">   28</span>&#160;{</div><div class="line"><a name="l00029"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a997e96288bdb106c922202e3f33d5d7b">   29</a></span>&#160;<span class="keyword">using</span> <a class="code" href="namespacearmnn.html#a997e96288bdb106c922202e3f33d5d7b">MinMaxRange</a> = std::pair&lt;float, float&gt;;</div><div class="line"><a name="l00030"></a><span class="lineno"><a class="line" href="namespacearmnn.html#ac757baefa4b72b54c38f713f86418f8a">   30</a></span>&#160;<span class="keyword">using</span> <a class="code" href="namespacearmnn.html#ac757baefa4b72b54c38f713f86418f8a">MinMaxRanges</a> = std::vector&lt;MinMaxRange&gt;;</div><div class="line"><a name="l00031"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a061aafb62b3769f55369845c3990ec7a">   31</a></span>&#160;<span class="keyword">using</span> <a class="code" href="namespacearmnn.html#a061aafb62b3769f55369845c3990ec7a">MinMaxRangeMap</a> = std::unordered_map&lt;LayerGuid, MinMaxRanges&gt;;</div><div class="line"><a name="l00032"></a><span class="lineno">   32</span>&#160;</div><div class="line"><a name="l00033"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">   33</a></span>&#160;<span class="keyword">const</span> <span class="keywordtype">float</span> <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a> = 255.0f;</div><div class="line"><a name="l00034"></a><span class="lineno">   34</span>&#160;<span class="comment">// Coinciding with calcution which for AsymmS8 which calculates scale on an unsigned basis</span></div><div class="line"><a name="l00035"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">   35</a></span>&#160;<span class="keyword">const</span> <span class="keywordtype">float</span> <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a> = 255.0f;</div><div class="line"><a name="l00036"></a><span class="lineno"><a class="line" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">   36</a></span>&#160;<span class="keyword">const</span> <span class="keywordtype">float</span> <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>  = 127.0f;</div><div class="line"><a name="l00037"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">   37</a></span>&#160;<span class="keyword">const</span> <span class="keywordtype">float</span> <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a> = 32767.0f;</div><div class="line"><a name="l00038"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a1a9a6dea47de10df8e3c76dd45df56fb">   38</a></span>&#160;<span class="keyword">const</span> <span class="keywordtype">float</span> <a class="code" href="namespacearmnn.html#a1a9a6dea47de10df8e3c76dd45df56fb">g_TestTolerance</a> = 0.000001f;</div><div class="line"><a name="l00039"></a><span class="lineno">   39</span>&#160;</div><div class="line"><a name="l00040"></a><span class="lineno">   40</span>&#160;<a class="code" href="_output_shape_of_squeeze_8cpp.html#ae3a6cb217a792718f2bd0e8f45e3ca9e">BOOST_AUTO_TEST_SUITE</a>(Quantizer)</div><div class="line"><a name="l00041"></a><span class="lineno">   41</span>&#160;</div><div class="line"><a name="l00042"></a><span class="lineno">   42</span>&#160;<span class="keyword">class </span>TestQuantization : <span class="keyword">public</span> <a class="code" href="classarmnn_1_1_layer_visitor_base.html">LayerVisitorBase</a>&lt;VisitorThrowingPolicy&gt;</div><div class="line"><a name="l00043"></a><span class="lineno">   43</span>&#160;{</div><div class="line"><a name="l00044"></a><span class="lineno">   44</span>&#160;<span class="keyword">public</span>:</div><div class="line"><a name="l00045"></a><span class="lineno">   45</span>&#160;    TestQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00046"></a><span class="lineno">   46</span>&#160;    : <a class="code" href="classarmnn_1_1_layer_visitor_base.html">LayerVisitorBase&lt;VisitorThrowingPolicy&gt;</a>()</div><div class="line"><a name="l00047"></a><span class="lineno">   47</span>&#160;    , m_InputShape(inputShape)</div><div class="line"><a name="l00048"></a><span class="lineno">   48</span>&#160;    , m_OutputShape(outputShape)</div><div class="line"><a name="l00049"></a><span class="lineno">   49</span>&#160;    , m_QuantizerOptions(<a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>()) {}</div><div class="line"><a name="l00050"></a><span class="lineno">   50</span>&#160;</div><div class="line"><a name="l00051"></a><span class="lineno">   51</span>&#160;    TestQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00052"></a><span class="lineno">   52</span>&#160;    : <a class="code" href="classarmnn_1_1_layer_visitor_base.html">LayerVisitorBase&lt;VisitorThrowingPolicy&gt;</a>()</div><div class="line"><a name="l00053"></a><span class="lineno">   53</span>&#160;    , m_InputShape(inputShape)</div><div class="line"><a name="l00054"></a><span class="lineno">   54</span>&#160;    , m_OutputShape(outputShape)</div><div class="line"><a name="l00055"></a><span class="lineno">   55</span>&#160;    , m_QuantizerOptions(options) {}</div><div class="line"><a name="l00056"></a><span class="lineno">   56</span>&#160;</div><div class="line"><a name="l00057"></a><span class="lineno">   57</span>&#160;    <span class="keywordtype">void</span> VisitInputLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l00058"></a><span class="lineno">   58</span>&#160;                         <a class="code" href="namespacearmnn.html#ab8cf8f9fb6792e654c2d8d8382f6f01b">LayerBindingId</a> <span class="keywordtype">id</span>,</div><div class="line"><a name="l00059"></a><span class="lineno">   59</span>&#160;                         <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l00060"></a><span class="lineno">   60</span>&#160;<span class="keyword">    </span>{</div><div class="line"><a name="l00061"></a><span class="lineno">   61</span>&#160;        boost::ignore_unused(<span class="keywordtype">id</span>, name);</div><div class="line"><a name="l00062"></a><span class="lineno">   62</span>&#160;        <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l00063"></a><span class="lineno">   63</span>&#160;        BOOST_TEST(m_InputShape == info.<a class="code" href="classarmnn_1_1_tensor_info.html#a8b5d0f8a24e9d9238f412260a552acf8">GetShape</a>());</div><div class="line"><a name="l00064"></a><span class="lineno">   64</span>&#160;        <span class="comment">// Based off current default [-15.0f, 15.0f]</span></div><div class="line"><a name="l00065"></a><span class="lineno">   65</span>&#160;        TestQuantizationParams(info, {30.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 128},</div><div class="line"><a name="l00066"></a><span class="lineno">   66</span>&#160;                                     {30.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>, 0},</div><div class="line"><a name="l00067"></a><span class="lineno">   67</span>&#160;                                     {15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a> , 0},</div><div class="line"><a name="l00068"></a><span class="lineno">   68</span>&#160;                                     {15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0});</div><div class="line"><a name="l00069"></a><span class="lineno">   69</span>&#160;    }</div><div class="line"><a name="l00070"></a><span class="lineno">   70</span>&#160;</div><div class="line"><a name="l00071"></a><span class="lineno">   71</span>&#160;    <span class="keywordtype">void</span> VisitOutputLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l00072"></a><span class="lineno">   72</span>&#160;                          <a class="code" href="namespacearmnn.html#ab8cf8f9fb6792e654c2d8d8382f6f01b">LayerBindingId</a> <span class="keywordtype">id</span>,</div><div class="line"><a name="l00073"></a><span class="lineno">   73</span>&#160;                          <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l00074"></a><span class="lineno">   74</span>&#160;<span class="keyword">    </span>{</div><div class="line"><a name="l00075"></a><span class="lineno">   75</span>&#160;        boost::ignore_unused(<span class="keywordtype">id</span>, name);</div><div class="line"><a name="l00076"></a><span class="lineno">   76</span>&#160;        <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_input_slot.html#a81fbf6103761e55061b62ba989b00f10">GetConnection</a>()-&gt;<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l00077"></a><span class="lineno">   77</span>&#160;        BOOST_TEST(m_OutputShape == info.<a class="code" href="classarmnn_1_1_tensor_info.html#a8b5d0f8a24e9d9238f412260a552acf8">GetShape</a>());</div><div class="line"><a name="l00078"></a><span class="lineno">   78</span>&#160;    }</div><div class="line"><a name="l00079"></a><span class="lineno">   79</span>&#160;</div><div class="line"><a name="l00080"></a><span class="lineno">   80</span>&#160;<span class="keyword">protected</span>:</div><div class="line"><a name="l00081"></a><span class="lineno">   81</span>&#160;    <span class="keywordtype">void</span> TestQuantizationParams(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>,</div><div class="line"><a name="l00082"></a><span class="lineno">   82</span>&#160;                                <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a>&amp; qAsymmU8Params,</div><div class="line"><a name="l00083"></a><span class="lineno">   83</span>&#160;                                <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a>&amp; qAsymmS8Params,</div><div class="line"><a name="l00084"></a><span class="lineno">   84</span>&#160;                                <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a>&amp; qSymmS8Params,</div><div class="line"><a name="l00085"></a><span class="lineno">   85</span>&#160;                                <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a>&amp; qSymmS16Params)</div><div class="line"><a name="l00086"></a><span class="lineno">   86</span>&#160;    {</div><div class="line"><a name="l00087"></a><span class="lineno">   87</span>&#160;        <span class="keywordflow">switch</span> (m_QuantizerOptions.m_ActivationFormat)</div><div class="line"><a name="l00088"></a><span class="lineno">   88</span>&#160;        {</div><div class="line"><a name="l00089"></a><span class="lineno">   89</span>&#160;            <span class="keywordflow">case</span> <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a0a3f57c876f5a230244c38e1453a8a6e">DataType::QAsymmU8</a>:</div><div class="line"><a name="l00090"></a><span class="lineno">   90</span>&#160;                TestQuantizationParamsImpl(</div><div class="line"><a name="l00091"></a><span class="lineno">   91</span>&#160;                    info, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a0a3f57c876f5a230244c38e1453a8a6e">DataType::QAsymmU8</a>, qAsymmU8Params.first, qAsymmU8Params.second);</div><div class="line"><a name="l00092"></a><span class="lineno">   92</span>&#160;                <span class="keywordflow">break</span>;</div><div class="line"><a name="l00093"></a><span class="lineno">   93</span>&#160;            <span class="keywordflow">case</span> <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>:</div><div class="line"><a name="l00094"></a><span class="lineno">   94</span>&#160;                TestQuantizationParamsImpl(</div><div class="line"><a name="l00095"></a><span class="lineno">   95</span>&#160;                    info, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>, qAsymmS8Params.first, qAsymmS8Params.second);</div><div class="line"><a name="l00096"></a><span class="lineno">   96</span>&#160;                <span class="keywordflow">break</span>;</div><div class="line"><a name="l00097"></a><span class="lineno">   97</span>&#160;            <span class="keywordflow">case</span> <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>:</div><div class="line"><a name="l00098"></a><span class="lineno">   98</span>&#160;                TestQuantizationParamsImpl(</div><div class="line"><a name="l00099"></a><span class="lineno">   99</span>&#160;                        info, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>, qSymmS8Params.first, qSymmS8Params.second);</div><div class="line"><a name="l00100"></a><span class="lineno">  100</span>&#160;                <span class="keywordflow">break</span>;</div><div class="line"><a name="l00101"></a><span class="lineno">  101</span>&#160;            <span class="keywordflow">case</span> <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>:</div><div class="line"><a name="l00102"></a><span class="lineno">  102</span>&#160;                TestQuantizationParamsImpl(</div><div class="line"><a name="l00103"></a><span class="lineno">  103</span>&#160;                    info, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>, qSymmS16Params.first, qSymmS16Params.second);</div><div class="line"><a name="l00104"></a><span class="lineno">  104</span>&#160;                <span class="keywordflow">break</span>;</div><div class="line"><a name="l00105"></a><span class="lineno">  105</span>&#160;            <span class="keywordflow">default</span>:</div><div class="line"><a name="l00106"></a><span class="lineno">  106</span>&#160;                <span class="keywordflow">throw</span> <a class="code" href="classarmnn_1_1_invalid_argument_exception.html">InvalidArgumentException</a>(<span class="stringliteral">&quot;Unsupported quantization target&quot;</span>);</div><div class="line"><a name="l00107"></a><span class="lineno">  107</span>&#160;        }</div><div class="line"><a name="l00108"></a><span class="lineno">  108</span>&#160;    }</div><div class="line"><a name="l00109"></a><span class="lineno">  109</span>&#160;</div><div class="line"><a name="l00110"></a><span class="lineno">  110</span>&#160;    <span class="keywordtype">void</span> TestDifferentQuantizationScale(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; info0, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; info1)</div><div class="line"><a name="l00111"></a><span class="lineno">  111</span>&#160;    {</div><div class="line"><a name="l00112"></a><span class="lineno">  112</span>&#160;        BOOST_TEST(info0.<a class="code" href="classarmnn_1_1_tensor_info.html#a047ca888c43bd7fb5702853bf72410d0">GetQuantizationScale</a>() != info1.<a class="code" href="classarmnn_1_1_tensor_info.html#a047ca888c43bd7fb5702853bf72410d0">GetQuantizationScale</a>());</div><div class="line"><a name="l00113"></a><span class="lineno">  113</span>&#160;    }</div><div class="line"><a name="l00114"></a><span class="lineno">  114</span>&#160;</div><div class="line"><a name="l00115"></a><span class="lineno">  115</span>&#160;    <span class="keywordtype">void</span> TestConstantQuantizationParams(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; info,</div><div class="line"><a name="l00116"></a><span class="lineno">  116</span>&#160;                                        <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a>&amp; params,</div><div class="line"><a name="l00117"></a><span class="lineno">  117</span>&#160;                                        <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6">DataType</a> dataType = <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a0a3f57c876f5a230244c38e1453a8a6e">DataType::QAsymmU8</a>)</div><div class="line"><a name="l00118"></a><span class="lineno">  118</span>&#160;    {</div><div class="line"><a name="l00119"></a><span class="lineno">  119</span>&#160;        boost::ignore_unused(dataType);</div><div class="line"><a name="l00120"></a><span class="lineno">  120</span>&#160;        TestQuantizationParamsImpl(info, dataType, params.first, params.second);</div><div class="line"><a name="l00121"></a><span class="lineno">  121</span>&#160;    }</div><div class="line"><a name="l00122"></a><span class="lineno">  122</span>&#160;</div><div class="line"><a name="l00123"></a><span class="lineno">  123</span>&#160;    <span class="keywordtype">void</span> TestBiasQuantizationParams(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; info,</div><div class="line"><a name="l00124"></a><span class="lineno">  124</span>&#160;                                    <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a>&amp; qAsymmU8Params,</div><div class="line"><a name="l00125"></a><span class="lineno">  125</span>&#160;                                    <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a>&amp; qAsymmS8Params,</div><div class="line"><a name="l00126"></a><span class="lineno">  126</span>&#160;                                    <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a>&amp; qSymmS8Params,</div><div class="line"><a name="l00127"></a><span class="lineno">  127</span>&#160;                                    <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a>&amp; qSymmS16Params,</div><div class="line"><a name="l00128"></a><span class="lineno">  128</span>&#160;                                    <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6">DataType</a> dataType = <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a0a3f57c876f5a230244c38e1453a8a6e">DataType::QAsymmU8</a>)</div><div class="line"><a name="l00129"></a><span class="lineno">  129</span>&#160;    {</div><div class="line"><a name="l00130"></a><span class="lineno">  130</span>&#160;        <span class="keywordflow">switch</span> (m_QuantizerOptions.m_ActivationFormat)</div><div class="line"><a name="l00131"></a><span class="lineno">  131</span>&#160;        {</div><div class="line"><a name="l00132"></a><span class="lineno">  132</span>&#160;            <span class="keywordflow">case</span> <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a0a3f57c876f5a230244c38e1453a8a6e">DataType::QAsymmU8</a>:</div><div class="line"><a name="l00133"></a><span class="lineno">  133</span>&#160;                TestQuantizationParamsImpl(info, dataType, qAsymmU8Params.first, qAsymmU8Params.second);</div><div class="line"><a name="l00134"></a><span class="lineno">  134</span>&#160;                <span class="keywordflow">break</span>;</div><div class="line"><a name="l00135"></a><span class="lineno">  135</span>&#160;            <span class="keywordflow">case</span> <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>:</div><div class="line"><a name="l00136"></a><span class="lineno">  136</span>&#160;                TestQuantizationParamsImpl(info, dataType, qAsymmS8Params.first, qAsymmS8Params.second);</div><div class="line"><a name="l00137"></a><span class="lineno">  137</span>&#160;                <span class="keywordflow">break</span>;</div><div class="line"><a name="l00138"></a><span class="lineno">  138</span>&#160;            <span class="keywordflow">case</span> <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>:</div><div class="line"><a name="l00139"></a><span class="lineno">  139</span>&#160;                TestQuantizationParamsImpl(info, dataType, qSymmS8Params.first, qSymmS8Params.second);</div><div class="line"><a name="l00140"></a><span class="lineno">  140</span>&#160;                <span class="keywordflow">break</span>;</div><div class="line"><a name="l00141"></a><span class="lineno">  141</span>&#160;            <span class="keywordflow">case</span> <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>:</div><div class="line"><a name="l00142"></a><span class="lineno">  142</span>&#160;                TestQuantizationParamsImpl(info, dataType, qSymmS16Params.first, qSymmS16Params.second);</div><div class="line"><a name="l00143"></a><span class="lineno">  143</span>&#160;                <span class="keywordflow">break</span>;</div><div class="line"><a name="l00144"></a><span class="lineno">  144</span>&#160;            <span class="keywordflow">default</span>:</div><div class="line"><a name="l00145"></a><span class="lineno">  145</span>&#160;                <span class="keywordflow">throw</span> <a class="code" href="classarmnn_1_1_invalid_argument_exception.html">InvalidArgumentException</a>(<span class="stringliteral">&quot;Unsupported quantization target&quot;</span>);</div><div class="line"><a name="l00146"></a><span class="lineno">  146</span>&#160;        }</div><div class="line"><a name="l00147"></a><span class="lineno">  147</span>&#160;    }</div><div class="line"><a name="l00148"></a><span class="lineno">  148</span>&#160;</div><div class="line"><a name="l00149"></a><span class="lineno">  149</span>&#160;    <span class="keywordtype">void</span> TestQuantizationOnLayersWithBiases(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l00150"></a><span class="lineno">  150</span>&#160;                                            <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a>&amp; weights,</div><div class="line"><a name="l00151"></a><span class="lineno">  151</span>&#160;                                            <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_optional.html">Optional&lt;ConstTensor&gt;</a>&amp; biases)</div><div class="line"><a name="l00152"></a><span class="lineno">  152</span>&#160;    {</div><div class="line"><a name="l00153"></a><span class="lineno">  153</span>&#160;        <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> info = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l00154"></a><span class="lineno">  154</span>&#160;        <span class="keywordtype">float</span> inputScaleQAsymmU8 = 30.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>;</div><div class="line"><a name="l00155"></a><span class="lineno">  155</span>&#160;        <span class="keywordtype">float</span> inputScaleQAsymmS8 = 30.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>;</div><div class="line"><a name="l00156"></a><span class="lineno">  156</span>&#160;        <span class="keywordtype">float</span> inputScaleQSymmS8  = 15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>;</div><div class="line"><a name="l00157"></a><span class="lineno">  157</span>&#160;        <span class="keywordtype">float</span> inputScaleQSymmS16 = 15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>;</div><div class="line"><a name="l00158"></a><span class="lineno">  158</span>&#160;        <span class="keywordtype">float</span> weightsScale       = 3.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>;</div><div class="line"><a name="l00159"></a><span class="lineno">  159</span>&#160;</div><div class="line"><a name="l00160"></a><span class="lineno">  160</span>&#160;        <span class="comment">// Based off default static range [-15.0f, 15.0f]</span></div><div class="line"><a name="l00161"></a><span class="lineno">  161</span>&#160;        TestQuantizationParams(info, {inputScaleQAsymmU8, 128},</div><div class="line"><a name="l00162"></a><span class="lineno">  162</span>&#160;                                     {inputScaleQAsymmS8, 0},</div><div class="line"><a name="l00163"></a><span class="lineno">  163</span>&#160;                                     {inputScaleQSymmS8, 0},</div><div class="line"><a name="l00164"></a><span class="lineno">  164</span>&#160;                                     {inputScaleQSymmS16, 0});</div><div class="line"><a name="l00165"></a><span class="lineno">  165</span>&#160;</div><div class="line"><a name="l00166"></a><span class="lineno">  166</span>&#160;        TestConstantQuantizationParams(weights.<a class="code" href="classarmnn_1_1_base_tensor.html#a8aeddebdcf02e1832b22203c08a6b678">GetInfo</a>(), {weightsScale, 85});</div><div class="line"><a name="l00167"></a><span class="lineno">  167</span>&#160;</div><div class="line"><a name="l00168"></a><span class="lineno">  168</span>&#160;        <span class="keywordflow">if</span> (biases.<a class="code" href="classarmnn_1_1_optional_base.html#a86b749ce2c4bc627fa8a1fcfaf0e314f">has_value</a>())</div><div class="line"><a name="l00169"></a><span class="lineno">  169</span>&#160;        {</div><div class="line"><a name="l00170"></a><span class="lineno">  170</span>&#160;            TestBiasQuantizationParams(biases.<a class="code" href="classarmnn_1_1_optional_reference_switch.html#a77c7d528ac063d870b8c8426ec81c1c3">value</a>().GetInfo(),</div><div class="line"><a name="l00171"></a><span class="lineno">  171</span>&#160;                                       {inputScaleQAsymmU8 * weightsScale, 0},</div><div class="line"><a name="l00172"></a><span class="lineno">  172</span>&#160;                                       {inputScaleQAsymmS8 * weightsScale, 0},</div><div class="line"><a name="l00173"></a><span class="lineno">  173</span>&#160;                                       {inputScaleQSymmS8  * weightsScale, 0},</div><div class="line"><a name="l00174"></a><span class="lineno">  174</span>&#160;                                       {inputScaleQSymmS16 * weightsScale, 0},</div><div class="line"><a name="l00175"></a><span class="lineno">  175</span>&#160;                                       <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6accedffbc6e5308e33d3843e8bdc0dad7">DataType::Signed32</a>);</div><div class="line"><a name="l00176"></a><span class="lineno">  176</span>&#160;        }</div><div class="line"><a name="l00177"></a><span class="lineno">  177</span>&#160;    }</div><div class="line"><a name="l00178"></a><span class="lineno">  178</span>&#160;</div><div class="line"><a name="l00179"></a><span class="lineno">  179</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> m_InputShape;</div><div class="line"><a name="l00180"></a><span class="lineno">  180</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> m_OutputShape;</div><div class="line"><a name="l00181"></a><span class="lineno">  181</span>&#160;</div><div class="line"><a name="l00182"></a><span class="lineno">  182</span>&#160;<span class="keyword">private</span>:</div><div class="line"><a name="l00183"></a><span class="lineno">  183</span>&#160;    <span class="keywordtype">void</span> TestQuantizationParamsImpl(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; info, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6">DataType</a> dataType, <span class="keywordtype">float</span> scale, int32_t offset)</div><div class="line"><a name="l00184"></a><span class="lineno">  184</span>&#160;    {</div><div class="line"><a name="l00185"></a><span class="lineno">  185</span>&#160;        BOOST_TEST((info.<a class="code" href="classarmnn_1_1_tensor_info.html#aea909c7327109228ef618d459015def3">GetDataType</a>() == dataType));</div><div class="line"><a name="l00186"></a><span class="lineno">  186</span>&#160;        BOOST_TEST(info.<a class="code" href="classarmnn_1_1_tensor_info.html#a770b51078da02f44a819e9f95d8058b5">GetQuantizationOffset</a>() == offset);</div><div class="line"><a name="l00187"></a><span class="lineno">  187</span>&#160;        BOOST_CHECK_CLOSE(info.<a class="code" href="classarmnn_1_1_tensor_info.html#a047ca888c43bd7fb5702853bf72410d0">GetQuantizationScale</a>(), scale, <a class="code" href="namespacearmnn.html#a1a9a6dea47de10df8e3c76dd45df56fb">g_TestTolerance</a>);</div><div class="line"><a name="l00188"></a><span class="lineno">  188</span>&#160;    }</div><div class="line"><a name="l00189"></a><span class="lineno">  189</span>&#160;</div><div class="line"><a name="l00190"></a><span class="lineno">  190</span>&#160;    <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> m_QuantizerOptions;</div><div class="line"><a name="l00191"></a><span class="lineno">  191</span>&#160;};</div><div class="line"><a name="l00192"></a><span class="lineno">  192</span>&#160;</div><div class="line"><a name="l00193"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">  193</a></span>&#160;<span class="keywordtype">void</span> <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_network.html">INetwork</a>* inputNetwork, <a class="code" href="classarmnn_1_1_i_layer_visitor.html">ILayerVisitor</a>&amp; visitor)</div><div class="line"><a name="l00194"></a><span class="lineno">  194</span>&#160;{</div><div class="line"><a name="l00195"></a><span class="lineno">  195</span>&#160;    <span class="keyword">auto</span> network = boost::polymorphic_downcast&lt;const Network*&gt;(inputNetwork);</div><div class="line"><a name="l00196"></a><span class="lineno">  196</span>&#160;    <span class="keyword">auto</span> graph = network-&gt;GetGraph().TopologicalSort();</div><div class="line"><a name="l00197"></a><span class="lineno">  197</span>&#160;</div><div class="line"><a name="l00198"></a><span class="lineno">  198</span>&#160;    <a class="code" href="namespacearmnn.html#a9835ef753dda5b5a2fe827680e41fda7">VisitLayers</a>(graph, visitor);</div><div class="line"><a name="l00199"></a><span class="lineno">  199</span>&#160;}</div><div class="line"><a name="l00200"></a><span class="lineno">  200</span>&#160;</div><div class="line"><a name="l00201"></a><span class="lineno">  201</span>&#160;<span class="keyword">class </span>TestAdditionQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l00202"></a><span class="lineno">  202</span>&#160;{</div><div class="line"><a name="l00203"></a><span class="lineno">  203</span>&#160;<span class="keyword">public</span>:</div><div class="line"><a name="l00204"></a><span class="lineno">  204</span>&#160;    TestAdditionQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00205"></a><span class="lineno">  205</span>&#160;    : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l00206"></a><span class="lineno">  206</span>&#160;</div><div class="line"><a name="l00207"></a><span class="lineno">  207</span>&#160;    TestAdditionQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l00208"></a><span class="lineno">  208</span>&#160;                             <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l00209"></a><span class="lineno">  209</span>&#160;                             <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00210"></a><span class="lineno">  210</span>&#160;    : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l00211"></a><span class="lineno">  211</span>&#160;</div><div class="line"><a name="l00212"></a><span class="lineno">  212</span>&#160;    <span class="keywordtype">void</span> VisitAdditionLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l00213"></a><span class="lineno">  213</span>&#160;                            <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l00214"></a><span class="lineno">  214</span>&#160;<span class="keyword">    </span>{</div><div class="line"><a name="l00215"></a><span class="lineno">  215</span>&#160;        boost::ignore_unused(name);</div><div class="line"><a name="l00216"></a><span class="lineno">  216</span>&#160;        <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l00217"></a><span class="lineno">  217</span>&#160;</div><div class="line"><a name="l00218"></a><span class="lineno">  218</span>&#160;        <span class="comment">// Based off default static range [-20.0f, 20.0f]</span></div><div class="line"><a name="l00219"></a><span class="lineno">  219</span>&#160;        TestQuantizationParams(info, {40.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 128},</div><div class="line"><a name="l00220"></a><span class="lineno">  220</span>&#160;                                     {40.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>, 0},</div><div class="line"><a name="l00221"></a><span class="lineno">  221</span>&#160;                                     {20.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0},</div><div class="line"><a name="l00222"></a><span class="lineno">  222</span>&#160;                                     {20.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0});</div><div class="line"><a name="l00223"></a><span class="lineno">  223</span>&#160;    }</div><div class="line"><a name="l00224"></a><span class="lineno">  224</span>&#160;};</div><div class="line"><a name="l00225"></a><span class="lineno">  225</span>&#160;</div><div class="line"><a name="l00226"></a><span class="lineno">  226</span>&#160;</div><div class="line"><a name="l00227"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a8baf97065d802063eb9bcdd1a066dc86">  227</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeAddition)</div><div class="line"><a name="l00228"></a><span class="lineno">  228</span>&#160;{</div><div class="line"><a name="l00229"></a><span class="lineno">  229</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l00230"></a><span class="lineno">  230</span>&#160;</div><div class="line"><a name="l00231"></a><span class="lineno">  231</span>&#160;    <span class="comment">// Add the layers</span></div><div class="line"><a name="l00232"></a><span class="lineno">  232</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input0 = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l00233"></a><span class="lineno">  233</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input1 = network-&gt;AddInputLayer(1);</div><div class="line"><a name="l00234"></a><span class="lineno">  234</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* addition = network-&gt;AddAdditionLayer();</div><div class="line"><a name="l00235"></a><span class="lineno">  235</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;AddOutputLayer(2);</div><div class="line"><a name="l00236"></a><span class="lineno">  236</span>&#160;</div><div class="line"><a name="l00237"></a><span class="lineno">  237</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l00238"></a><span class="lineno">  238</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l00239"></a><span class="lineno">  239</span>&#160;    input1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(1));</div><div class="line"><a name="l00240"></a><span class="lineno">  240</span>&#160;    addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l00241"></a><span class="lineno">  241</span>&#160;</div><div class="line"><a name="l00242"></a><span class="lineno">  242</span>&#160;    <span class="comment">// Set TensorInfo</span></div><div class="line"><a name="l00243"></a><span class="lineno">  243</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l00244"></a><span class="lineno">  244</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l00245"></a><span class="lineno">  245</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l00246"></a><span class="lineno">  246</span>&#160;    input1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l00247"></a><span class="lineno">  247</span>&#160;    addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l00248"></a><span class="lineno">  248</span>&#160;</div><div class="line"><a name="l00249"></a><span class="lineno">  249</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmU8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a0a3f57c876f5a230244c38e1453a8a6e">DataType::QAsymmU8</a>);</div><div class="line"><a name="l00250"></a><span class="lineno">  250</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmU8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00251"></a><span class="lineno">  251</span>&#160;    TestAdditionQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l00252"></a><span class="lineno">  252</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l00253"></a><span class="lineno">  253</span>&#160;</div><div class="line"><a name="l00254"></a><span class="lineno">  254</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l00255"></a><span class="lineno">  255</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00256"></a><span class="lineno">  256</span>&#160;    TestAdditionQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l00257"></a><span class="lineno">  257</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l00258"></a><span class="lineno">  258</span>&#160;</div><div class="line"><a name="l00259"></a><span class="lineno">  259</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l00260"></a><span class="lineno">  260</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00261"></a><span class="lineno">  261</span>&#160;    TestAdditionQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l00262"></a><span class="lineno">  262</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l00263"></a><span class="lineno">  263</span>&#160;</div><div class="line"><a name="l00264"></a><span class="lineno">  264</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l00265"></a><span class="lineno">  265</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l00266"></a><span class="lineno">  266</span>&#160;    TestAdditionQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l00267"></a><span class="lineno">  267</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l00268"></a><span class="lineno">  268</span>&#160;}</div><div class="line"><a name="l00269"></a><span class="lineno">  269</span>&#160;</div><div class="line"><a name="l00270"></a><span class="lineno">  270</span>&#160;<span class="keyword">class </span>TestActivationQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l00271"></a><span class="lineno">  271</span>&#160;{</div><div class="line"><a name="l00272"></a><span class="lineno">  272</span>&#160;<span class="keyword">public</span>:</div><div class="line"><a name="l00273"></a><span class="lineno">  273</span>&#160;    TestActivationQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00274"></a><span class="lineno">  274</span>&#160;    : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l00275"></a><span class="lineno">  275</span>&#160;</div><div class="line"><a name="l00276"></a><span class="lineno">  276</span>&#160;    TestActivationQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l00277"></a><span class="lineno">  277</span>&#160;                               <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l00278"></a><span class="lineno">  278</span>&#160;                               <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00279"></a><span class="lineno">  279</span>&#160;    : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l00280"></a><span class="lineno">  280</span>&#160;</div><div class="line"><a name="l00281"></a><span class="lineno">  281</span>&#160;    <span class="keywordtype">void</span> VisitActivationLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l00282"></a><span class="lineno">  282</span>&#160;                              <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_activation_descriptor.html">ActivationDescriptor</a>&amp; descriptor,</div><div class="line"><a name="l00283"></a><span class="lineno">  283</span>&#160;                              <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l00284"></a><span class="lineno">  284</span>&#160;<span class="keyword">    </span>{</div><div class="line"><a name="l00285"></a><span class="lineno">  285</span>&#160;        boost::ignore_unused(descriptor, name);</div><div class="line"><a name="l00286"></a><span class="lineno">  286</span>&#160;</div><div class="line"><a name="l00287"></a><span class="lineno">  287</span>&#160;        <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l00288"></a><span class="lineno">  288</span>&#160;</div><div class="line"><a name="l00289"></a><span class="lineno">  289</span>&#160;        <span class="comment">// Based off default static range [0.0f, 15.0f]</span></div><div class="line"><a name="l00290"></a><span class="lineno">  290</span>&#160;        TestQuantizationParams(info, {15.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 0},</div><div class="line"><a name="l00291"></a><span class="lineno">  291</span>&#160;                                     {15.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>, -128},</div><div class="line"><a name="l00292"></a><span class="lineno">  292</span>&#160;                                     {15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>, 0},</div><div class="line"><a name="l00293"></a><span class="lineno">  293</span>&#160;                                     {15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0});</div><div class="line"><a name="l00294"></a><span class="lineno">  294</span>&#160;    }</div><div class="line"><a name="l00295"></a><span class="lineno">  295</span>&#160;};</div><div class="line"><a name="l00296"></a><span class="lineno">  296</span>&#160;</div><div class="line"><a name="l00297"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a5fbc1479db5f4ff70a750cf02d58971b">  297</a></span>&#160;<a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> <a class="code" href="namespacearmnn.html#a5fbc1479db5f4ff70a750cf02d58971b">CreateNetworkWithActivationLayer</a>(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_activation_descriptor.html">ActivationDescriptor</a>&amp; descriptor, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; shape)</div><div class="line"><a name="l00298"></a><span class="lineno">  298</span>&#160;{</div><div class="line"><a name="l00299"></a><span class="lineno">  299</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l00300"></a><span class="lineno">  300</span>&#160;</div><div class="line"><a name="l00301"></a><span class="lineno">  301</span>&#160;    <span class="comment">// Add the layers</span></div><div class="line"><a name="l00302"></a><span class="lineno">  302</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input0 = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l00303"></a><span class="lineno">  303</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* activation = network-&gt;AddActivationLayer(descriptor);</div><div class="line"><a name="l00304"></a><span class="lineno">  304</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;AddOutputLayer(2);</div><div class="line"><a name="l00305"></a><span class="lineno">  305</span>&#160;</div><div class="line"><a name="l00306"></a><span class="lineno">  306</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l00307"></a><span class="lineno">  307</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(activation-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l00308"></a><span class="lineno">  308</span>&#160;    activation-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l00309"></a><span class="lineno">  309</span>&#160;</div><div class="line"><a name="l00310"></a><span class="lineno">  310</span>&#160;    <span class="comment">// Set TensorInfo</span></div><div class="line"><a name="l00311"></a><span class="lineno">  311</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l00312"></a><span class="lineno">  312</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l00313"></a><span class="lineno">  313</span>&#160;    activation-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l00314"></a><span class="lineno">  314</span>&#160;</div><div class="line"><a name="l00315"></a><span class="lineno">  315</span>&#160;    <span class="keywordflow">return</span> network;</div><div class="line"><a name="l00316"></a><span class="lineno">  316</span>&#160;}</div><div class="line"><a name="l00317"></a><span class="lineno">  317</span>&#160;</div><div class="line"><a name="l00318"></a><span class="lineno"><a class="line" href="namespacearmnn.html#aa9c6c1a7b5380a99a536f4740f87dd59">  318</a></span>&#160;<a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> <a class="code" href="namespacearmnn.html#aa9c6c1a7b5380a99a536f4740f87dd59">CreateNetworkWithInputOutputLayers</a>()</div><div class="line"><a name="l00319"></a><span class="lineno">  319</span>&#160;{</div><div class="line"><a name="l00320"></a><span class="lineno">  320</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l00321"></a><span class="lineno">  321</span>&#160;</div><div class="line"><a name="l00322"></a><span class="lineno">  322</span>&#160;    <span class="comment">// Add input/output layers</span></div><div class="line"><a name="l00323"></a><span class="lineno">  323</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* inputLayer = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l00324"></a><span class="lineno">  324</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;AddOutputLayer(1);</div><div class="line"><a name="l00325"></a><span class="lineno">  325</span>&#160;</div><div class="line"><a name="l00326"></a><span class="lineno">  326</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l00327"></a><span class="lineno">  327</span>&#160;    inputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l00328"></a><span class="lineno">  328</span>&#160;</div><div class="line"><a name="l00329"></a><span class="lineno">  329</span>&#160;    <span class="comment">// Set TensorInfo</span></div><div class="line"><a name="l00330"></a><span class="lineno">  330</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{8U};</div><div class="line"><a name="l00331"></a><span class="lineno">  331</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l00332"></a><span class="lineno">  332</span>&#160;    inputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l00333"></a><span class="lineno">  333</span>&#160;</div><div class="line"><a name="l00334"></a><span class="lineno">  334</span>&#160;    <span class="keywordflow">return</span> network;</div><div class="line"><a name="l00335"></a><span class="lineno">  335</span>&#160;}</div><div class="line"><a name="l00336"></a><span class="lineno">  336</span>&#160;</div><div class="line"><a name="l00337"></a><span class="lineno"><a class="line" href="namespacearmnn.html#ae52296dff1f4879854f320d59f92574e">  337</a></span>&#160;<a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#ae52296dff1f4879854f320d59f92574e">GetInputTensorInfo</a>(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_network.html">Network</a>* network)</div><div class="line"><a name="l00338"></a><span class="lineno">  338</span>&#160;{</div><div class="line"><a name="l00339"></a><span class="lineno">  339</span>&#160;    <span class="keywordflow">for</span> (<span class="keyword">auto</span>&amp;&amp; inputLayer : network-&gt;<a class="code" href="classarmnn_1_1_network.html#afe0a4f719f9752a405e71878da7012ba">GetGraph</a>().<a class="code" href="classarmnn_1_1_graph.html#a919fb58873ef3a6549e4490e226f2eae">GetInputLayers</a>())</div><div class="line"><a name="l00340"></a><span class="lineno">  340</span>&#160;    {</div><div class="line"><a name="l00341"></a><span class="lineno">  341</span>&#160;        BOOST_ASSERT_MSG(inputLayer-&gt;GetNumOutputSlots() == 1, <span class="stringliteral">&quot;Input layer should have exactly 1 output slot&quot;</span>);</div><div class="line"><a name="l00342"></a><span class="lineno">  342</span>&#160;        <span class="keywordflow">return</span> inputLayer-&gt;GetOutputSlot(0).GetTensorInfo();</div><div class="line"><a name="l00343"></a><span class="lineno">  343</span>&#160;    }</div><div class="line"><a name="l00344"></a><span class="lineno">  344</span>&#160;    <span class="keywordflow">throw</span> <a class="code" href="classarmnn_1_1_invalid_argument_exception.html">InvalidArgumentException</a>(<span class="stringliteral">&quot;Network has no input layers&quot;</span>);</div><div class="line"><a name="l00345"></a><span class="lineno">  345</span>&#160;}</div><div class="line"><a name="l00346"></a><span class="lineno">  346</span>&#160;</div><div class="line"><a name="l00347"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a9cec088786b209989fe9e04e1be9636d">  347</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(InputOutputLayerDynamicQuant)</div><div class="line"><a name="l00348"></a><span class="lineno">  348</span>&#160;{</div><div class="line"><a name="l00349"></a><span class="lineno">  349</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="namespacearmnn.html#aa9c6c1a7b5380a99a536f4740f87dd59">CreateNetworkWithInputOutputLayers</a>();</div><div class="line"><a name="l00350"></a><span class="lineno">  350</span>&#160;</div><div class="line"><a name="l00351"></a><span class="lineno">  351</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">armnn::TensorInfo</a> tensorInfo = <a class="code" href="namespacearmnn.html#ae52296dff1f4879854f320d59f92574e">GetInputTensorInfo</a>(boost::polymorphic_downcast&lt;const Network*&gt;(network.get()));</div><div class="line"><a name="l00352"></a><span class="lineno">  352</span>&#160;</div><div class="line"><a name="l00353"></a><span class="lineno">  353</span>&#160;    <span class="comment">// Outliers -56 and 98</span></div><div class="line"><a name="l00354"></a><span class="lineno">  354</span>&#160;    std::vector&lt;float&gt; inputData({0, 0, 0, -56, 98, 0, 0, 0});</div><div class="line"><a name="l00355"></a><span class="lineno">  355</span>&#160;    <a class="code" href="classarmnn_1_1_const_tensor.html">armnn::ConstTensor</a> inputTensor(tensorInfo, inputData.data());</div><div class="line"><a name="l00356"></a><span class="lineno">  356</span>&#160;</div><div class="line"><a name="l00357"></a><span class="lineno">  357</span>&#160;    <a class="code" href="namespacearmnn.html#aa01bce88f89975a5a031db4cc8861527">InputTensors</a> inputTensors;</div><div class="line"><a name="l00358"></a><span class="lineno">  358</span>&#160;    inputTensors.push_back(std::make_pair(0, inputTensor));</div><div class="line"><a name="l00359"></a><span class="lineno">  359</span>&#160;</div><div class="line"><a name="l00360"></a><span class="lineno">  360</span>&#160;    <a class="code" href="namespacearmnn.html#a41119e261eec9343888d2ceab1e4999a">armnn::INetworkQuantizerPtr</a> quantizer = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">armnn::INetworkQuantizer::Create</a>(network.get());</div><div class="line"><a name="l00361"></a><span class="lineno">  361</span>&#160;</div><div class="line"><a name="l00362"></a><span class="lineno">  362</span>&#160;    quantizer-&gt;Refine(inputTensors);</div><div class="line"><a name="l00363"></a><span class="lineno">  363</span>&#160;</div><div class="line"><a name="l00364"></a><span class="lineno">  364</span>&#160;    <span class="comment">// Outliers -77 and 65</span></div><div class="line"><a name="l00365"></a><span class="lineno">  365</span>&#160;    std::vector&lt;float&gt; inputData2({0, -77, 0, -56, 65, 0, 0, 0});</div><div class="line"><a name="l00366"></a><span class="lineno">  366</span>&#160;    <a class="code" href="classarmnn_1_1_const_tensor.html">armnn::ConstTensor</a> inputTensor2(tensorInfo, inputData2.data());</div><div class="line"><a name="l00367"></a><span class="lineno">  367</span>&#160;    <a class="code" href="namespacearmnn.html#aa01bce88f89975a5a031db4cc8861527">InputTensors</a> inputTensors2;</div><div class="line"><a name="l00368"></a><span class="lineno">  368</span>&#160;    inputTensors2.push_back(std::make_pair(0, inputTensor2));</div><div class="line"><a name="l00369"></a><span class="lineno">  369</span>&#160;</div><div class="line"><a name="l00370"></a><span class="lineno">  370</span>&#160;    quantizer-&gt;Refine(inputTensors2);</div><div class="line"><a name="l00371"></a><span class="lineno">  371</span>&#160;</div><div class="line"><a name="l00372"></a><span class="lineno">  372</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetwork = quantizer-&gt;ExportNetwork();</div><div class="line"><a name="l00373"></a><span class="lineno">  373</span>&#160;    <span class="comment">// Output Layer should be quantized for a min max of -77 and 98</span></div><div class="line"><a name="l00374"></a><span class="lineno">  374</span>&#160;    <span class="comment">// according to QU8 Quantization Scheme</span></div><div class="line"><a name="l00375"></a><span class="lineno">  375</span>&#160;    std::unique_ptr&lt;IQuantizationScheme&gt; quantizationScheme = std::make_unique&lt;QAsymmU8QuantizationScheme&gt;();</div><div class="line"><a name="l00376"></a><span class="lineno">  376</span>&#160;    <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qParams = quantizationScheme-&gt;ComputeScheme(-77.0, 98.0);</div><div class="line"><a name="l00377"></a><span class="lineno">  377</span>&#160;</div><div class="line"><a name="l00378"></a><span class="lineno">  378</span>&#160;    <span class="keyword">class </span><a class="code" href="classarmnn_1_1_test_output_layer_visitor.html">TestOutputLayerVisitor</a> : <span class="keyword">public</span> <a class="code" href="classarmnn_1_1_layer_visitor_base.html">LayerVisitorBase</a>&lt;VisitorNoThrowPolicy&gt;</div><div class="line"><a name="l00379"></a><span class="lineno">  379</span>&#160;    {</div><div class="line"><a name="l00380"></a><span class="lineno">  380</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l00381"></a><span class="lineno">  381</span>&#160;        <a class="code" href="classarmnn_1_1_test_output_layer_visitor.html">TestOutputLayerVisitor</a>(<span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a>&amp; offsetScalePair, <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6">DataType</a>&amp; dataType) :</div><div class="line"><a name="l00382"></a><span class="lineno">  382</span>&#160;            m_OffsetScalePair(offsetScalePair), m_DataType(dataType) {}</div><div class="line"><a name="l00383"></a><span class="lineno">  383</span>&#160;</div><div class="line"><a name="l00384"></a><span class="lineno">  384</span>&#160;        <span class="keywordtype">void</span> VisitOutputLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l00385"></a><span class="lineno">  385</span>&#160;                                      <a class="code" href="namespacearmnn.html#ab8cf8f9fb6792e654c2d8d8382f6f01b">LayerBindingId</a> <span class="keywordtype">id</span>,</div><div class="line"><a name="l00386"></a><span class="lineno">  386</span>&#160;                                      <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l00387"></a><span class="lineno">  387</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l00388"></a><span class="lineno">  388</span>&#160;            boost::ignore_unused(<span class="keywordtype">id</span>, name);</div><div class="line"><a name="l00389"></a><span class="lineno">  389</span>&#160;            <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_input_slot.html#a81fbf6103761e55061b62ba989b00f10">GetConnection</a>()-&gt;<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l00390"></a><span class="lineno">  390</span>&#160;            BOOST_CHECK_MESSAGE(info.<a class="code" href="classarmnn_1_1_tensor_info.html#aea909c7327109228ef618d459015def3">GetDataType</a>() == m_DataType,</div><div class="line"><a name="l00391"></a><span class="lineno">  391</span>&#160;                                std::string(<a class="code" href="namespacearmnn.html#a81b5ff8545adad19a1c9d4ca076d552c">armnn::GetDataTypeName</a>(info.<a class="code" href="classarmnn_1_1_tensor_info.html#aea909c7327109228ef618d459015def3">GetDataType</a>()))</div><div class="line"><a name="l00392"></a><span class="lineno">  392</span>&#160;                                        .append(<span class="stringliteral">&quot; == &quot;</span>).append(<a class="code" href="namespacearmnn.html#a81b5ff8545adad19a1c9d4ca076d552c">armnn::GetDataTypeName</a>(m_DataType)));</div><div class="line"><a name="l00393"></a><span class="lineno">  393</span>&#160;            <span class="comment">// int_32t</span></div><div class="line"><a name="l00394"></a><span class="lineno">  394</span>&#160;            <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a0c262ba6f6c189a2d092d127c1b7627b">BOOST_CHECK</a>(info.<a class="code" href="classarmnn_1_1_tensor_info.html#a770b51078da02f44a819e9f95d8058b5">GetQuantizationOffset</a>() == m_OffsetScalePair.second);</div><div class="line"><a name="l00395"></a><span class="lineno">  395</span>&#160;            <span class="comment">// float</span></div><div class="line"><a name="l00396"></a><span class="lineno">  396</span>&#160;            BOOST_TEST(info.<a class="code" href="classarmnn_1_1_tensor_info.html#a047ca888c43bd7fb5702853bf72410d0">GetQuantizationScale</a>() == m_OffsetScalePair.first, boost::test_tools::tolerance(0.001));</div><div class="line"><a name="l00397"></a><span class="lineno">  397</span>&#160;        }</div><div class="line"><a name="l00398"></a><span class="lineno">  398</span>&#160;</div><div class="line"><a name="l00399"></a><span class="lineno">  399</span>&#160;    <span class="keyword">private</span>:</div><div class="line"><a name="l00400"></a><span class="lineno">  400</span>&#160;        <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> m_OffsetScalePair;</div><div class="line"><a name="l00401"></a><span class="lineno">  401</span>&#160;        <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6">DataType</a> m_DataType;</div><div class="line"><a name="l00402"></a><span class="lineno">  402</span>&#160;    };</div><div class="line"><a name="l00403"></a><span class="lineno">  403</span>&#160;</div><div class="line"><a name="l00404"></a><span class="lineno">  404</span>&#160;    <a class="code" href="classarmnn_1_1_test_output_layer_visitor.html">TestOutputLayerVisitor</a> visitor(qParams, quantizationScheme-&gt;GetDataType());</div><div class="line"><a name="l00405"></a><span class="lineno">  405</span>&#160;    quantizedNetwork-&gt;Accept(visitor);</div><div class="line"><a name="l00406"></a><span class="lineno">  406</span>&#160;}</div><div class="line"><a name="l00407"></a><span class="lineno">  407</span>&#160;</div><div class="line"><a name="l00408"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a7db6a78bb6eedbea7f0525f1fe59de28">  408</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeAbsActivation)</div><div class="line"><a name="l00409"></a><span class="lineno">  409</span>&#160;{</div><div class="line"><a name="l00410"></a><span class="lineno">  410</span>&#160;    <a class="code" href="structarmnn_1_1_activation_descriptor.html">ActivationDescriptor</a> descriptor;</div><div class="line"><a name="l00411"></a><span class="lineno">  411</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#af10fa7883e3579950f477bee92a64844">m_Function</a> = <a class="code" href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaa1e34af023adeb7d5f484f8eb4b9826b6">ActivationFunction::Abs</a>;</div><div class="line"><a name="l00412"></a><span class="lineno">  412</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a017b2990003a014234f13e999dc7c689">m_A</a>        = 3.5f;</div><div class="line"><a name="l00413"></a><span class="lineno">  413</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a28c4c9cb15f6be3499abbc46b356060b">m_B</a>        = -10.0f;</div><div class="line"><a name="l00414"></a><span class="lineno">  414</span>&#160;</div><div class="line"><a name="l00415"></a><span class="lineno">  415</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l00416"></a><span class="lineno">  416</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="namespacearmnn.html#a5fbc1479db5f4ff70a750cf02d58971b">CreateNetworkWithActivationLayer</a>(descriptor, shape);</div><div class="line"><a name="l00417"></a><span class="lineno">  417</span>&#160;</div><div class="line"><a name="l00418"></a><span class="lineno">  418</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmU8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a0a3f57c876f5a230244c38e1453a8a6e">DataType::QAsymmU8</a>);</div><div class="line"><a name="l00419"></a><span class="lineno">  419</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmU8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00420"></a><span class="lineno">  420</span>&#160;    TestActivationQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l00421"></a><span class="lineno">  421</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l00422"></a><span class="lineno">  422</span>&#160;</div><div class="line"><a name="l00423"></a><span class="lineno">  423</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l00424"></a><span class="lineno">  424</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00425"></a><span class="lineno">  425</span>&#160;    TestActivationQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l00426"></a><span class="lineno">  426</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l00427"></a><span class="lineno">  427</span>&#160;</div><div class="line"><a name="l00428"></a><span class="lineno">  428</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l00429"></a><span class="lineno">  429</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00430"></a><span class="lineno">  430</span>&#160;    TestActivationQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l00431"></a><span class="lineno">  431</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l00432"></a><span class="lineno">  432</span>&#160;</div><div class="line"><a name="l00433"></a><span class="lineno">  433</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l00434"></a><span class="lineno">  434</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l00435"></a><span class="lineno">  435</span>&#160;    TestActivationQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l00436"></a><span class="lineno">  436</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l00437"></a><span class="lineno">  437</span>&#160;}</div><div class="line"><a name="l00438"></a><span class="lineno">  438</span>&#160;</div><div class="line"><a name="l00439"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a2df3b432de50a9b9e8b486aa53e11cc5">  439</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeLinearActivation)</div><div class="line"><a name="l00440"></a><span class="lineno">  440</span>&#160;{</div><div class="line"><a name="l00441"></a><span class="lineno">  441</span>&#160;    <a class="code" href="structarmnn_1_1_activation_descriptor.html">ActivationDescriptor</a> descriptor;</div><div class="line"><a name="l00442"></a><span class="lineno">  442</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#af10fa7883e3579950f477bee92a64844">m_Function</a> = <a class="code" href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaa32a843da6ea40ab3b17a3421ccdf671b">ActivationFunction::Linear</a>;</div><div class="line"><a name="l00443"></a><span class="lineno">  443</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a017b2990003a014234f13e999dc7c689">m_A</a>        = 3.5f;</div><div class="line"><a name="l00444"></a><span class="lineno">  444</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a28c4c9cb15f6be3499abbc46b356060b">m_B</a>        = -10.0f;</div><div class="line"><a name="l00445"></a><span class="lineno">  445</span>&#160;</div><div class="line"><a name="l00446"></a><span class="lineno">  446</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l00447"></a><span class="lineno">  447</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="namespacearmnn.html#a5fbc1479db5f4ff70a750cf02d58971b">CreateNetworkWithActivationLayer</a>(descriptor, shape);</div><div class="line"><a name="l00448"></a><span class="lineno">  448</span>&#160;</div><div class="line"><a name="l00449"></a><span class="lineno">  449</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l00450"></a><span class="lineno">  450</span>&#160;    TestActivationQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l00451"></a><span class="lineno">  451</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l00452"></a><span class="lineno">  452</span>&#160;</div><div class="line"><a name="l00453"></a><span class="lineno">  453</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l00454"></a><span class="lineno">  454</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00455"></a><span class="lineno">  455</span>&#160;    TestActivationQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l00456"></a><span class="lineno">  456</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l00457"></a><span class="lineno">  457</span>&#160;</div><div class="line"><a name="l00458"></a><span class="lineno">  458</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l00459"></a><span class="lineno">  459</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00460"></a><span class="lineno">  460</span>&#160;    TestActivationQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l00461"></a><span class="lineno">  461</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l00462"></a><span class="lineno">  462</span>&#160;</div><div class="line"><a name="l00463"></a><span class="lineno">  463</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l00464"></a><span class="lineno">  464</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l00465"></a><span class="lineno">  465</span>&#160;    TestActivationQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l00466"></a><span class="lineno">  466</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l00467"></a><span class="lineno">  467</span>&#160;}</div><div class="line"><a name="l00468"></a><span class="lineno">  468</span>&#160;</div><div class="line"><a name="l00469"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a3dd219b394b8186d1849ee595193268d">  469</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeReLuActivation)</div><div class="line"><a name="l00470"></a><span class="lineno">  470</span>&#160;{</div><div class="line"><a name="l00471"></a><span class="lineno">  471</span>&#160;    <a class="code" href="structarmnn_1_1_activation_descriptor.html">ActivationDescriptor</a> descriptor;</div><div class="line"><a name="l00472"></a><span class="lineno">  472</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#af10fa7883e3579950f477bee92a64844">m_Function</a> = <a class="code" href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaa3d90c0a5ab3fcf8e6f6608cb3d3a1559">ActivationFunction::ReLu</a>;</div><div class="line"><a name="l00473"></a><span class="lineno">  473</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a017b2990003a014234f13e999dc7c689">m_A</a>        = 3.5f;</div><div class="line"><a name="l00474"></a><span class="lineno">  474</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a28c4c9cb15f6be3499abbc46b356060b">m_B</a>        = -10.0f;</div><div class="line"><a name="l00475"></a><span class="lineno">  475</span>&#160;</div><div class="line"><a name="l00476"></a><span class="lineno">  476</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l00477"></a><span class="lineno">  477</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="namespacearmnn.html#a5fbc1479db5f4ff70a750cf02d58971b">CreateNetworkWithActivationLayer</a>(descriptor, shape);</div><div class="line"><a name="l00478"></a><span class="lineno">  478</span>&#160;</div><div class="line"><a name="l00479"></a><span class="lineno">  479</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l00480"></a><span class="lineno">  480</span>&#160;    TestActivationQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l00481"></a><span class="lineno">  481</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l00482"></a><span class="lineno">  482</span>&#160;</div><div class="line"><a name="l00483"></a><span class="lineno">  483</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l00484"></a><span class="lineno">  484</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00485"></a><span class="lineno">  485</span>&#160;    TestActivationQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l00486"></a><span class="lineno">  486</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l00487"></a><span class="lineno">  487</span>&#160;</div><div class="line"><a name="l00488"></a><span class="lineno">  488</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l00489"></a><span class="lineno">  489</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00490"></a><span class="lineno">  490</span>&#160;    TestActivationQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l00491"></a><span class="lineno">  491</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l00492"></a><span class="lineno">  492</span>&#160;</div><div class="line"><a name="l00493"></a><span class="lineno">  493</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l00494"></a><span class="lineno">  494</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l00495"></a><span class="lineno">  495</span>&#160;    TestActivationQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l00496"></a><span class="lineno">  496</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l00497"></a><span class="lineno">  497</span>&#160;}</div><div class="line"><a name="l00498"></a><span class="lineno">  498</span>&#160;</div><div class="line"><a name="l00499"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a52e948b4bffc16a3933d812dbc384833">  499</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeSoftReLuActivation)</div><div class="line"><a name="l00500"></a><span class="lineno">  500</span>&#160;{</div><div class="line"><a name="l00501"></a><span class="lineno">  501</span>&#160;    <a class="code" href="structarmnn_1_1_activation_descriptor.html">ActivationDescriptor</a> descriptor;</div><div class="line"><a name="l00502"></a><span class="lineno">  502</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#af10fa7883e3579950f477bee92a64844">m_Function</a> = <a class="code" href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaa6bba7052636d1740303b1b2ef3b53fef">ActivationFunction::SoftReLu</a>;</div><div class="line"><a name="l00503"></a><span class="lineno">  503</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a017b2990003a014234f13e999dc7c689">m_A</a>        = 3.5f;</div><div class="line"><a name="l00504"></a><span class="lineno">  504</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a28c4c9cb15f6be3499abbc46b356060b">m_B</a>        = -10.0f;</div><div class="line"><a name="l00505"></a><span class="lineno">  505</span>&#160;</div><div class="line"><a name="l00506"></a><span class="lineno">  506</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l00507"></a><span class="lineno">  507</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="namespacearmnn.html#a5fbc1479db5f4ff70a750cf02d58971b">CreateNetworkWithActivationLayer</a>(descriptor, shape);</div><div class="line"><a name="l00508"></a><span class="lineno">  508</span>&#160;</div><div class="line"><a name="l00509"></a><span class="lineno">  509</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l00510"></a><span class="lineno">  510</span>&#160;    TestActivationQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l00511"></a><span class="lineno">  511</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l00512"></a><span class="lineno">  512</span>&#160;</div><div class="line"><a name="l00513"></a><span class="lineno">  513</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l00514"></a><span class="lineno">  514</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00515"></a><span class="lineno">  515</span>&#160;    TestActivationQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l00516"></a><span class="lineno">  516</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l00517"></a><span class="lineno">  517</span>&#160;</div><div class="line"><a name="l00518"></a><span class="lineno">  518</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l00519"></a><span class="lineno">  519</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00520"></a><span class="lineno">  520</span>&#160;    TestActivationQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l00521"></a><span class="lineno">  521</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l00522"></a><span class="lineno">  522</span>&#160;</div><div class="line"><a name="l00523"></a><span class="lineno">  523</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l00524"></a><span class="lineno">  524</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l00525"></a><span class="lineno">  525</span>&#160;    TestActivationQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l00526"></a><span class="lineno">  526</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l00527"></a><span class="lineno">  527</span>&#160;}</div><div class="line"><a name="l00528"></a><span class="lineno">  528</span>&#160;</div><div class="line"><a name="l00529"></a><span class="lineno"><a class="line" href="namespacearmnn.html#abf109580225cb949565c8223bceadd5d">  529</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeBoundedReluActivation)</div><div class="line"><a name="l00530"></a><span class="lineno">  530</span>&#160;{</div><div class="line"><a name="l00531"></a><span class="lineno">  531</span>&#160;    <span class="keyword">class </span>TestBoundedReluActivationQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l00532"></a><span class="lineno">  532</span>&#160;    {</div><div class="line"><a name="l00533"></a><span class="lineno">  533</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l00534"></a><span class="lineno">  534</span>&#160;        TestBoundedReluActivationQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00535"></a><span class="lineno">  535</span>&#160;        : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l00536"></a><span class="lineno">  536</span>&#160;</div><div class="line"><a name="l00537"></a><span class="lineno">  537</span>&#160;        TestBoundedReluActivationQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l00538"></a><span class="lineno">  538</span>&#160;                                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l00539"></a><span class="lineno">  539</span>&#160;                                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00540"></a><span class="lineno">  540</span>&#160;        : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l00541"></a><span class="lineno">  541</span>&#160;</div><div class="line"><a name="l00542"></a><span class="lineno">  542</span>&#160;        <span class="keywordtype">void</span> VisitActivationLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l00543"></a><span class="lineno">  543</span>&#160;                                  <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_activation_descriptor.html">ActivationDescriptor</a>&amp; descriptor,</div><div class="line"><a name="l00544"></a><span class="lineno">  544</span>&#160;                                  <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l00545"></a><span class="lineno">  545</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l00546"></a><span class="lineno">  546</span>&#160;            boost::ignore_unused(descriptor, name);</div><div class="line"><a name="l00547"></a><span class="lineno">  547</span>&#160;            <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l00548"></a><span class="lineno">  548</span>&#160;</div><div class="line"><a name="l00549"></a><span class="lineno">  549</span>&#160;            <span class="comment">// Based off default static range [0.0f, 3.5f]</span></div><div class="line"><a name="l00550"></a><span class="lineno">  550</span>&#160;            TestQuantizationParams(info, {3.5f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 0},</div><div class="line"><a name="l00551"></a><span class="lineno">  551</span>&#160;                                         {3.5f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>, -128},</div><div class="line"><a name="l00552"></a><span class="lineno">  552</span>&#160;                                         {3.5f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0},</div><div class="line"><a name="l00553"></a><span class="lineno">  553</span>&#160;                                         {3.5f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0});</div><div class="line"><a name="l00554"></a><span class="lineno">  554</span>&#160;        }</div><div class="line"><a name="l00555"></a><span class="lineno">  555</span>&#160;    };</div><div class="line"><a name="l00556"></a><span class="lineno">  556</span>&#160;</div><div class="line"><a name="l00557"></a><span class="lineno">  557</span>&#160;    <a class="code" href="structarmnn_1_1_activation_descriptor.html">ActivationDescriptor</a> descriptor;</div><div class="line"><a name="l00558"></a><span class="lineno">  558</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#af10fa7883e3579950f477bee92a64844">m_Function</a> = <a class="code" href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaabc5a0f0d6e7cf7fca73299dcd46c0f0d">ActivationFunction::BoundedReLu</a>;</div><div class="line"><a name="l00559"></a><span class="lineno">  559</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a017b2990003a014234f13e999dc7c689">m_A</a>        = 3.5f;</div><div class="line"><a name="l00560"></a><span class="lineno">  560</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a28c4c9cb15f6be3499abbc46b356060b">m_B</a>        = -10.0f;</div><div class="line"><a name="l00561"></a><span class="lineno">  561</span>&#160;</div><div class="line"><a name="l00562"></a><span class="lineno">  562</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l00563"></a><span class="lineno">  563</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="namespacearmnn.html#a5fbc1479db5f4ff70a750cf02d58971b">CreateNetworkWithActivationLayer</a>(descriptor, shape);</div><div class="line"><a name="l00564"></a><span class="lineno">  564</span>&#160;</div><div class="line"><a name="l00565"></a><span class="lineno">  565</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l00566"></a><span class="lineno">  566</span>&#160;    TestBoundedReluActivationQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l00567"></a><span class="lineno">  567</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l00568"></a><span class="lineno">  568</span>&#160;</div><div class="line"><a name="l00569"></a><span class="lineno">  569</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l00570"></a><span class="lineno">  570</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00571"></a><span class="lineno">  571</span>&#160;    TestBoundedReluActivationQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l00572"></a><span class="lineno">  572</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l00573"></a><span class="lineno">  573</span>&#160;</div><div class="line"><a name="l00574"></a><span class="lineno">  574</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l00575"></a><span class="lineno">  575</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00576"></a><span class="lineno">  576</span>&#160;    TestBoundedReluActivationQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l00577"></a><span class="lineno">  577</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l00578"></a><span class="lineno">  578</span>&#160;</div><div class="line"><a name="l00579"></a><span class="lineno">  579</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l00580"></a><span class="lineno">  580</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l00581"></a><span class="lineno">  581</span>&#160;    TestBoundedReluActivationQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l00582"></a><span class="lineno">  582</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l00583"></a><span class="lineno">  583</span>&#160;}</div><div class="line"><a name="l00584"></a><span class="lineno">  584</span>&#160;</div><div class="line"><a name="l00585"></a><span class="lineno"><a class="line" href="namespacearmnn.html#acbf871a6ec0726bfe2746e761a278108">  585</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeTanHActivation)</div><div class="line"><a name="l00586"></a><span class="lineno">  586</span>&#160;{</div><div class="line"><a name="l00587"></a><span class="lineno">  587</span>&#160;    <span class="keyword">class </span>TestTanHActivationQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l00588"></a><span class="lineno">  588</span>&#160;    {</div><div class="line"><a name="l00589"></a><span class="lineno">  589</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l00590"></a><span class="lineno">  590</span>&#160;        TestTanHActivationQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00591"></a><span class="lineno">  591</span>&#160;        : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l00592"></a><span class="lineno">  592</span>&#160;</div><div class="line"><a name="l00593"></a><span class="lineno">  593</span>&#160;        TestTanHActivationQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l00594"></a><span class="lineno">  594</span>&#160;                                       <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l00595"></a><span class="lineno">  595</span>&#160;                                       <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00596"></a><span class="lineno">  596</span>&#160;        : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l00597"></a><span class="lineno">  597</span>&#160;</div><div class="line"><a name="l00598"></a><span class="lineno">  598</span>&#160;        <span class="keywordtype">void</span> VisitActivationLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l00599"></a><span class="lineno">  599</span>&#160;                                  <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_activation_descriptor.html">ActivationDescriptor</a>&amp; descriptor,</div><div class="line"><a name="l00600"></a><span class="lineno">  600</span>&#160;                                  <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l00601"></a><span class="lineno">  601</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l00602"></a><span class="lineno">  602</span>&#160;            boost::ignore_unused(descriptor, name);</div><div class="line"><a name="l00603"></a><span class="lineno">  603</span>&#160;            <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l00604"></a><span class="lineno">  604</span>&#160;</div><div class="line"><a name="l00605"></a><span class="lineno">  605</span>&#160;            <span class="comment">// Based off default static range [-1.0f, 1.0f]</span></div><div class="line"><a name="l00606"></a><span class="lineno">  606</span>&#160;            TestQuantizationParams(</div><div class="line"><a name="l00607"></a><span class="lineno">  607</span>&#160;                info, {2.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 128},</div><div class="line"><a name="l00608"></a><span class="lineno">  608</span>&#160;                      {2.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>,   0},</div><div class="line"><a name="l00609"></a><span class="lineno">  609</span>&#160;                      {1.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a> ,   0},</div><div class="line"><a name="l00610"></a><span class="lineno">  610</span>&#160;                      {1.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>,   0});</div><div class="line"><a name="l00611"></a><span class="lineno">  611</span>&#160;        }</div><div class="line"><a name="l00612"></a><span class="lineno">  612</span>&#160;    };</div><div class="line"><a name="l00613"></a><span class="lineno">  613</span>&#160;</div><div class="line"><a name="l00614"></a><span class="lineno">  614</span>&#160;    <a class="code" href="structarmnn_1_1_activation_descriptor.html">ActivationDescriptor</a> descriptor;</div><div class="line"><a name="l00615"></a><span class="lineno">  615</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#af10fa7883e3579950f477bee92a64844">m_Function</a> = <a class="code" href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaa23b68da1de2b77d74da9da2635722a3e">ActivationFunction::TanH</a>;</div><div class="line"><a name="l00616"></a><span class="lineno">  616</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a017b2990003a014234f13e999dc7c689">m_A</a>        = 3.5f;</div><div class="line"><a name="l00617"></a><span class="lineno">  617</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a28c4c9cb15f6be3499abbc46b356060b">m_B</a>        = -10.0f;</div><div class="line"><a name="l00618"></a><span class="lineno">  618</span>&#160;</div><div class="line"><a name="l00619"></a><span class="lineno">  619</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l00620"></a><span class="lineno">  620</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="namespacearmnn.html#a5fbc1479db5f4ff70a750cf02d58971b">CreateNetworkWithActivationLayer</a>(descriptor, shape);</div><div class="line"><a name="l00621"></a><span class="lineno">  621</span>&#160;</div><div class="line"><a name="l00622"></a><span class="lineno">  622</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l00623"></a><span class="lineno">  623</span>&#160;    TestTanHActivationQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l00624"></a><span class="lineno">  624</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l00625"></a><span class="lineno">  625</span>&#160;</div><div class="line"><a name="l00626"></a><span class="lineno">  626</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l00627"></a><span class="lineno">  627</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00628"></a><span class="lineno">  628</span>&#160;    TestTanHActivationQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l00629"></a><span class="lineno">  629</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l00630"></a><span class="lineno">  630</span>&#160;</div><div class="line"><a name="l00631"></a><span class="lineno">  631</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l00632"></a><span class="lineno">  632</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00633"></a><span class="lineno">  633</span>&#160;    TestTanHActivationQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l00634"></a><span class="lineno">  634</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l00635"></a><span class="lineno">  635</span>&#160;</div><div class="line"><a name="l00636"></a><span class="lineno">  636</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l00637"></a><span class="lineno">  637</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l00638"></a><span class="lineno">  638</span>&#160;    TestTanHActivationQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l00639"></a><span class="lineno">  639</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l00640"></a><span class="lineno">  640</span>&#160;}</div><div class="line"><a name="l00641"></a><span class="lineno">  641</span>&#160;</div><div class="line"><a name="l00642"></a><span class="lineno">  642</span>&#160;<span class="keyword">class </span>TestLeakyReLuActivationQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l00643"></a><span class="lineno">  643</span>&#160;{</div><div class="line"><a name="l00644"></a><span class="lineno">  644</span>&#160;<span class="keyword">public</span>:</div><div class="line"><a name="l00645"></a><span class="lineno">  645</span>&#160;    TestLeakyReLuActivationQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00646"></a><span class="lineno">  646</span>&#160;    : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l00647"></a><span class="lineno">  647</span>&#160;</div><div class="line"><a name="l00648"></a><span class="lineno">  648</span>&#160;    TestLeakyReLuActivationQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l00649"></a><span class="lineno">  649</span>&#160;                                        <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l00650"></a><span class="lineno">  650</span>&#160;                                        <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00651"></a><span class="lineno">  651</span>&#160;    : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l00652"></a><span class="lineno">  652</span>&#160;</div><div class="line"><a name="l00653"></a><span class="lineno">  653</span>&#160;    <span class="keywordtype">void</span> VisitActivationLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l00654"></a><span class="lineno">  654</span>&#160;                              <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_activation_descriptor.html">ActivationDescriptor</a>&amp; descriptor,</div><div class="line"><a name="l00655"></a><span class="lineno">  655</span>&#160;                              <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l00656"></a><span class="lineno">  656</span>&#160;<span class="keyword">    </span>{</div><div class="line"><a name="l00657"></a><span class="lineno">  657</span>&#160;        boost::ignore_unused(descriptor, name);</div><div class="line"><a name="l00658"></a><span class="lineno">  658</span>&#160;        <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l00659"></a><span class="lineno">  659</span>&#160;</div><div class="line"><a name="l00660"></a><span class="lineno">  660</span>&#160;        <span class="comment">// Based off default static range [-5.0f, 15.0f]</span></div><div class="line"><a name="l00661"></a><span class="lineno">  661</span>&#160;        TestQuantizationParams(info, {20.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 64},</div><div class="line"><a name="l00662"></a><span class="lineno">  662</span>&#160;                                     {20.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>,-64},</div><div class="line"><a name="l00663"></a><span class="lineno">  663</span>&#160;                                     {15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a> ,  0},</div><div class="line"><a name="l00664"></a><span class="lineno">  664</span>&#160;                                     {15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>,  0});</div><div class="line"><a name="l00665"></a><span class="lineno">  665</span>&#160;    }</div><div class="line"><a name="l00666"></a><span class="lineno">  666</span>&#160;</div><div class="line"><a name="l00667"></a><span class="lineno">  667</span>&#160;<span class="keyword">protected</span>:</div><div class="line"><a name="l00668"></a><span class="lineno">  668</span>&#160;    <span class="comment">// Used by the descendant classes which test layers</span></div><div class="line"><a name="l00669"></a><span class="lineno">  669</span>&#160;    <span class="comment">// that are forwarding their parent layer settings</span></div><div class="line"><a name="l00670"></a><span class="lineno">  670</span>&#160;    <span class="keywordtype">void</span> CheckForwardedQuantizationSettings(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer)</div><div class="line"><a name="l00671"></a><span class="lineno">  671</span>&#160;    {</div><div class="line"><a name="l00672"></a><span class="lineno">  672</span>&#160;        <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l00673"></a><span class="lineno">  673</span>&#160;        TestQuantizationParams(info, {20.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 64},</div><div class="line"><a name="l00674"></a><span class="lineno">  674</span>&#160;                                     {20.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>,-64},</div><div class="line"><a name="l00675"></a><span class="lineno">  675</span>&#160;                                     {15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,   0},</div><div class="line"><a name="l00676"></a><span class="lineno">  676</span>&#160;                                     {15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>,  0});</div><div class="line"><a name="l00677"></a><span class="lineno">  677</span>&#160;    }</div><div class="line"><a name="l00678"></a><span class="lineno">  678</span>&#160;};</div><div class="line"><a name="l00679"></a><span class="lineno">  679</span>&#160;</div><div class="line"><a name="l00680"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a32068047cc7b37f1bed1830508891526">  680</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeLeakyReLuActivation)</div><div class="line"><a name="l00681"></a><span class="lineno">  681</span>&#160;{</div><div class="line"><a name="l00682"></a><span class="lineno">  682</span>&#160;    <a class="code" href="structarmnn_1_1_activation_descriptor.html">ActivationDescriptor</a> descriptor;</div><div class="line"><a name="l00683"></a><span class="lineno">  683</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#af10fa7883e3579950f477bee92a64844">m_Function</a> = <a class="code" href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaacb7667f5ec2f6e8a5943b781ba6c2735">ActivationFunction::LeakyReLu</a>;</div><div class="line"><a name="l00684"></a><span class="lineno">  684</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a017b2990003a014234f13e999dc7c689">m_A</a>        = 3.5f;</div><div class="line"><a name="l00685"></a><span class="lineno">  685</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a28c4c9cb15f6be3499abbc46b356060b">m_B</a>        = -10.0f;</div><div class="line"><a name="l00686"></a><span class="lineno">  686</span>&#160;</div><div class="line"><a name="l00687"></a><span class="lineno">  687</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l00688"></a><span class="lineno">  688</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="namespacearmnn.html#a5fbc1479db5f4ff70a750cf02d58971b">CreateNetworkWithActivationLayer</a>(descriptor, shape);</div><div class="line"><a name="l00689"></a><span class="lineno">  689</span>&#160;</div><div class="line"><a name="l00690"></a><span class="lineno">  690</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l00691"></a><span class="lineno">  691</span>&#160;    TestLeakyReLuActivationQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l00692"></a><span class="lineno">  692</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l00693"></a><span class="lineno">  693</span>&#160;</div><div class="line"><a name="l00694"></a><span class="lineno">  694</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l00695"></a><span class="lineno">  695</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00696"></a><span class="lineno">  696</span>&#160;    TestLeakyReLuActivationQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l00697"></a><span class="lineno">  697</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l00698"></a><span class="lineno">  698</span>&#160;</div><div class="line"><a name="l00699"></a><span class="lineno">  699</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l00700"></a><span class="lineno">  700</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00701"></a><span class="lineno">  701</span>&#160;    TestLeakyReLuActivationQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l00702"></a><span class="lineno">  702</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l00703"></a><span class="lineno">  703</span>&#160;</div><div class="line"><a name="l00704"></a><span class="lineno">  704</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l00705"></a><span class="lineno">  705</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l00706"></a><span class="lineno">  706</span>&#160;    TestLeakyReLuActivationQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l00707"></a><span class="lineno">  707</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l00708"></a><span class="lineno">  708</span>&#160;}</div><div class="line"><a name="l00709"></a><span class="lineno">  709</span>&#160;</div><div class="line"><a name="l00710"></a><span class="lineno"><a class="line" href="namespacearmnn.html#adf59f87645d301e9b56dd70aed350e54">  710</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeBatchNorm)</div><div class="line"><a name="l00711"></a><span class="lineno">  711</span>&#160;{</div><div class="line"><a name="l00712"></a><span class="lineno">  712</span>&#160;    <span class="keyword">class </span>TestBatchNormalizationQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l00713"></a><span class="lineno">  713</span>&#160;    {</div><div class="line"><a name="l00714"></a><span class="lineno">  714</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l00715"></a><span class="lineno">  715</span>&#160;        TestBatchNormalizationQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00716"></a><span class="lineno">  716</span>&#160;        : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l00717"></a><span class="lineno">  717</span>&#160;</div><div class="line"><a name="l00718"></a><span class="lineno">  718</span>&#160;        TestBatchNormalizationQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l00719"></a><span class="lineno">  719</span>&#160;                                           <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l00720"></a><span class="lineno">  720</span>&#160;                                           <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00721"></a><span class="lineno">  721</span>&#160;        : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l00722"></a><span class="lineno">  722</span>&#160;</div><div class="line"><a name="l00723"></a><span class="lineno">  723</span>&#160;        <span class="keywordtype">void</span> VisitBatchNormalizationLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l00724"></a><span class="lineno">  724</span>&#160;                                          <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_batch_normalization_descriptor.html">BatchNormalizationDescriptor</a>&amp; desc,</div><div class="line"><a name="l00725"></a><span class="lineno">  725</span>&#160;                                          <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a>&amp; mean,</div><div class="line"><a name="l00726"></a><span class="lineno">  726</span>&#160;                                          <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a>&amp; variance,</div><div class="line"><a name="l00727"></a><span class="lineno">  727</span>&#160;                                          <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a>&amp; beta,</div><div class="line"><a name="l00728"></a><span class="lineno">  728</span>&#160;                                          <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a>&amp; gamma,</div><div class="line"><a name="l00729"></a><span class="lineno">  729</span>&#160;                                          <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l00730"></a><span class="lineno">  730</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l00731"></a><span class="lineno">  731</span>&#160;            boost::ignore_unused(desc, name);</div><div class="line"><a name="l00732"></a><span class="lineno">  732</span>&#160;            <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l00733"></a><span class="lineno">  733</span>&#160;</div><div class="line"><a name="l00734"></a><span class="lineno">  734</span>&#160;            <span class="comment">// Based off default static range [-15.0f, 15.0f]</span></div><div class="line"><a name="l00735"></a><span class="lineno">  735</span>&#160;            TestQuantizationParams(</div><div class="line"><a name="l00736"></a><span class="lineno">  736</span>&#160;                info, {30.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 128},</div><div class="line"><a name="l00737"></a><span class="lineno">  737</span>&#160;                      {30.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>,  0},</div><div class="line"><a name="l00738"></a><span class="lineno">  738</span>&#160;                      {15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0},</div><div class="line"><a name="l00739"></a><span class="lineno">  739</span>&#160;                      {15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0});</div><div class="line"><a name="l00740"></a><span class="lineno">  740</span>&#160;</div><div class="line"><a name="l00741"></a><span class="lineno">  741</span>&#160;            <span class="comment">// Test constants</span></div><div class="line"><a name="l00742"></a><span class="lineno">  742</span>&#160;            TestConstantQuantizationParams(mean.<a class="code" href="classarmnn_1_1_base_tensor.html#a8aeddebdcf02e1832b22203c08a6b678">GetInfo</a>(), {3.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 85});</div><div class="line"><a name="l00743"></a><span class="lineno">  743</span>&#160;            TestConstantQuantizationParams(variance.<a class="code" href="classarmnn_1_1_base_tensor.html#a8aeddebdcf02e1832b22203c08a6b678">GetInfo</a>(), {3.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 85});</div><div class="line"><a name="l00744"></a><span class="lineno">  744</span>&#160;            TestConstantQuantizationParams(beta.<a class="code" href="classarmnn_1_1_base_tensor.html#a8aeddebdcf02e1832b22203c08a6b678">GetInfo</a>(), {3.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 85});</div><div class="line"><a name="l00745"></a><span class="lineno">  745</span>&#160;            TestConstantQuantizationParams(gamma.<a class="code" href="classarmnn_1_1_base_tensor.html#a8aeddebdcf02e1832b22203c08a6b678">GetInfo</a>(), {3.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 85});</div><div class="line"><a name="l00746"></a><span class="lineno">  746</span>&#160;        }</div><div class="line"><a name="l00747"></a><span class="lineno">  747</span>&#160;    };</div><div class="line"><a name="l00748"></a><span class="lineno">  748</span>&#160;</div><div class="line"><a name="l00749"></a><span class="lineno">  749</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l00750"></a><span class="lineno">  750</span>&#160;</div><div class="line"><a name="l00751"></a><span class="lineno">  751</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{3U};</div><div class="line"><a name="l00752"></a><span class="lineno">  752</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l00753"></a><span class="lineno">  753</span>&#160;</div><div class="line"><a name="l00754"></a><span class="lineno">  754</span>&#160;    std::vector&lt;float&gt; meanData{-1.0f, 1.5f, 2.0f};</div><div class="line"><a name="l00755"></a><span class="lineno">  755</span>&#160;    std::vector&lt;float&gt; varData{-1.0f, 1.5f, 2.0f};</div><div class="line"><a name="l00756"></a><span class="lineno">  756</span>&#160;    std::vector&lt;float&gt; betaData{-1.0f, 1.5f, 2.0f};</div><div class="line"><a name="l00757"></a><span class="lineno">  757</span>&#160;    std::vector&lt;float&gt; gammaData{-1.0f, 1.5f, 2.0f};</div><div class="line"><a name="l00758"></a><span class="lineno">  758</span>&#160;</div><div class="line"><a name="l00759"></a><span class="lineno">  759</span>&#160;    <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a> mean(info, meanData);</div><div class="line"><a name="l00760"></a><span class="lineno">  760</span>&#160;    <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a> var(info, varData);</div><div class="line"><a name="l00761"></a><span class="lineno">  761</span>&#160;    <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a> beta(info, betaData);</div><div class="line"><a name="l00762"></a><span class="lineno">  762</span>&#160;    <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a> gamma(info, gammaData);</div><div class="line"><a name="l00763"></a><span class="lineno">  763</span>&#160;</div><div class="line"><a name="l00764"></a><span class="lineno">  764</span>&#160;    <a class="code" href="structarmnn_1_1_batch_normalization_descriptor.html">BatchNormalizationDescriptor</a> desc;</div><div class="line"><a name="l00765"></a><span class="lineno">  765</span>&#160;</div><div class="line"><a name="l00766"></a><span class="lineno">  766</span>&#160;    <span class="comment">// Add the layers</span></div><div class="line"><a name="l00767"></a><span class="lineno">  767</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input0 = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l00768"></a><span class="lineno">  768</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* batchNorm = network-&gt;AddBatchNormalizationLayer(desc, mean, var, beta, gamma);</div><div class="line"><a name="l00769"></a><span class="lineno">  769</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;AddOutputLayer(1);</div><div class="line"><a name="l00770"></a><span class="lineno">  770</span>&#160;</div><div class="line"><a name="l00771"></a><span class="lineno">  771</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l00772"></a><span class="lineno">  772</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(batchNorm-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l00773"></a><span class="lineno">  773</span>&#160;    batchNorm-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l00774"></a><span class="lineno">  774</span>&#160;</div><div class="line"><a name="l00775"></a><span class="lineno">  775</span>&#160;    <span class="comment">// Set TensorInfo</span></div><div class="line"><a name="l00776"></a><span class="lineno">  776</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l00777"></a><span class="lineno">  777</span>&#160;    batchNorm-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l00778"></a><span class="lineno">  778</span>&#160;</div><div class="line"><a name="l00779"></a><span class="lineno">  779</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l00780"></a><span class="lineno">  780</span>&#160;    TestBatchNormalizationQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l00781"></a><span class="lineno">  781</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l00782"></a><span class="lineno">  782</span>&#160;</div><div class="line"><a name="l00783"></a><span class="lineno">  783</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l00784"></a><span class="lineno">  784</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00785"></a><span class="lineno">  785</span>&#160;    TestBatchNormalizationQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l00786"></a><span class="lineno">  786</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l00787"></a><span class="lineno">  787</span>&#160;</div><div class="line"><a name="l00788"></a><span class="lineno">  788</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l00789"></a><span class="lineno">  789</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00790"></a><span class="lineno">  790</span>&#160;    TestBatchNormalizationQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l00791"></a><span class="lineno">  791</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l00792"></a><span class="lineno">  792</span>&#160;</div><div class="line"><a name="l00793"></a><span class="lineno">  793</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> QQsymm16Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l00794"></a><span class="lineno">  794</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), QQsymm16Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00795"></a><span class="lineno">  795</span>&#160;    TestBatchNormalizationQuantization validatorQSymmS16(QQsymm16Options, shape, shape);</div><div class="line"><a name="l00796"></a><span class="lineno">  796</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l00797"></a><span class="lineno">  797</span>&#160;}</div><div class="line"><a name="l00798"></a><span class="lineno">  798</span>&#160;</div><div class="line"><a name="l00799"></a><span class="lineno"><a class="line" href="namespacearmnn.html#ae91bc23bf56bb5f9c2e0ddb1fc7be75e">  799</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeDepthToSpace)</div><div class="line"><a name="l00800"></a><span class="lineno">  800</span>&#160;{</div><div class="line"><a name="l00801"></a><span class="lineno">  801</span>&#160;    <span class="keyword">class </span>TestDepthToSpaceQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l00802"></a><span class="lineno">  802</span>&#160;    {</div><div class="line"><a name="l00803"></a><span class="lineno">  803</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l00804"></a><span class="lineno">  804</span>&#160;        TestDepthToSpaceQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00805"></a><span class="lineno">  805</span>&#160;            : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l00806"></a><span class="lineno">  806</span>&#160;</div><div class="line"><a name="l00807"></a><span class="lineno">  807</span>&#160;        TestDepthToSpaceQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l00808"></a><span class="lineno">  808</span>&#160;                                     <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l00809"></a><span class="lineno">  809</span>&#160;                                     <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00810"></a><span class="lineno">  810</span>&#160;            : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l00811"></a><span class="lineno">  811</span>&#160;</div><div class="line"><a name="l00812"></a><span class="lineno">  812</span>&#160;        <span class="keyword">virtual</span> <span class="keywordtype">void</span> VisitDepthToSpaceLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l00813"></a><span class="lineno">  813</span>&#160;                                            <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_space_to_depth_descriptor.html">DepthToSpaceDescriptor</a>&amp; desc,</div><div class="line"><a name="l00814"></a><span class="lineno">  814</span>&#160;                                            <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)</div><div class="line"><a name="l00815"></a><span class="lineno">  815</span>&#160;        {</div><div class="line"><a name="l00816"></a><span class="lineno">  816</span>&#160;            boost::ignore_unused(desc, name);</div><div class="line"><a name="l00817"></a><span class="lineno">  817</span>&#160;            <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l00818"></a><span class="lineno">  818</span>&#160;</div><div class="line"><a name="l00819"></a><span class="lineno">  819</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qAsymmU8Params{ 30.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 128 };</div><div class="line"><a name="l00820"></a><span class="lineno">  820</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qAsymmS8Params{ 30.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>, 0 };</div><div class="line"><a name="l00821"></a><span class="lineno">  821</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qSymmS8Params { 15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0 };</div><div class="line"><a name="l00822"></a><span class="lineno">  822</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qSymmS16Params{ 15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0 };</div><div class="line"><a name="l00823"></a><span class="lineno">  823</span>&#160;</div><div class="line"><a name="l00824"></a><span class="lineno">  824</span>&#160;            TestQuantizationParams(info, qAsymmU8Params, qAsymmS8Params, qSymmS8Params, qSymmS16Params);</div><div class="line"><a name="l00825"></a><span class="lineno">  825</span>&#160;        }</div><div class="line"><a name="l00826"></a><span class="lineno">  826</span>&#160;    };</div><div class="line"><a name="l00827"></a><span class="lineno">  827</span>&#160;</div><div class="line"><a name="l00828"></a><span class="lineno">  828</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> inputShape { 1, 2, 2, 4 };</div><div class="line"><a name="l00829"></a><span class="lineno">  829</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> outputShape{ 1, 4, 4, 1 };</div><div class="line"><a name="l00830"></a><span class="lineno">  830</span>&#160;</div><div class="line"><a name="l00831"></a><span class="lineno">  831</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> inputInfo (inputShape,  <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l00832"></a><span class="lineno">  832</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> outputInfo(outputShape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l00833"></a><span class="lineno">  833</span>&#160;</div><div class="line"><a name="l00834"></a><span class="lineno">  834</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l00835"></a><span class="lineno">  835</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_space_to_depth_descriptor.html">DepthToSpaceDescriptor</a> descriptor(2, <a class="code" href="namespacearmnn.html#ad1d5cce2d9e9a5d61c243e5c989112e0ad066db54b89b0912e7e7c6da51e2da51">armnn::DataLayout::NHWC</a>);</div><div class="line"><a name="l00836"></a><span class="lineno">  836</span>&#160;</div><div class="line"><a name="l00837"></a><span class="lineno">  837</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* inputLayer        = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l00838"></a><span class="lineno">  838</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* depthToSpaceLayer = network-&gt;AddDepthToSpaceLayer(descriptor);</div><div class="line"><a name="l00839"></a><span class="lineno">  839</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* outputLayer       = network-&gt;AddOutputLayer(0);</div><div class="line"><a name="l00840"></a><span class="lineno">  840</span>&#160;</div><div class="line"><a name="l00841"></a><span class="lineno">  841</span>&#160;    inputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(depthToSpaceLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l00842"></a><span class="lineno">  842</span>&#160;    depthToSpaceLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(outputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l00843"></a><span class="lineno">  843</span>&#160;</div><div class="line"><a name="l00844"></a><span class="lineno">  844</span>&#160;    inputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(inputInfo);</div><div class="line"><a name="l00845"></a><span class="lineno">  845</span>&#160;    depthToSpaceLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(outputInfo);</div><div class="line"><a name="l00846"></a><span class="lineno">  846</span>&#160;</div><div class="line"><a name="l00847"></a><span class="lineno">  847</span>&#160;    <span class="comment">// test QAsymmU8 quantization</span></div><div class="line"><a name="l00848"></a><span class="lineno">  848</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l00849"></a><span class="lineno">  849</span>&#160;    TestDepthToSpaceQuantization validatorQAsymmU8(inputShape, outputShape);</div><div class="line"><a name="l00850"></a><span class="lineno">  850</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l00851"></a><span class="lineno">  851</span>&#160;</div><div class="line"><a name="l00852"></a><span class="lineno">  852</span>&#160;    <span class="comment">// test QAsymmS8 quantization</span></div><div class="line"><a name="l00853"></a><span class="lineno">  853</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l00854"></a><span class="lineno">  854</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00855"></a><span class="lineno">  855</span>&#160;    TestDepthToSpaceQuantization validatorQAsymmS8(qAsymmS8Options, inputShape, outputShape);</div><div class="line"><a name="l00856"></a><span class="lineno">  856</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l00857"></a><span class="lineno">  857</span>&#160;</div><div class="line"><a name="l00858"></a><span class="lineno">  858</span>&#160;    <span class="comment">// test QSymmS8 quantization</span></div><div class="line"><a name="l00859"></a><span class="lineno">  859</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l00860"></a><span class="lineno">  860</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00861"></a><span class="lineno">  861</span>&#160;    TestDepthToSpaceQuantization validatorQSymmS8(qSymmS8Options, inputShape, outputShape);</div><div class="line"><a name="l00862"></a><span class="lineno">  862</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l00863"></a><span class="lineno">  863</span>&#160;</div><div class="line"><a name="l00864"></a><span class="lineno">  864</span>&#160;    <span class="comment">// test QSymmS16 quantization</span></div><div class="line"><a name="l00865"></a><span class="lineno">  865</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> Qsymm16Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l00866"></a><span class="lineno">  866</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), Qsymm16Options)-&gt;ExportNetwork();</div><div class="line"><a name="l00867"></a><span class="lineno">  867</span>&#160;    TestDepthToSpaceQuantization validatorQSymmS16(Qsymm16Options, inputShape, outputShape);</div><div class="line"><a name="l00868"></a><span class="lineno">  868</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l00869"></a><span class="lineno">  869</span>&#160;}</div><div class="line"><a name="l00870"></a><span class="lineno">  870</span>&#160;</div><div class="line"><a name="l00871"></a><span class="lineno"><a class="line" href="namespacearmnn.html#aa6281ed3090b74167170c8f692688de5">  871</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(OverrideInputRangeEmptyNetwork)</div><div class="line"><a name="l00872"></a><span class="lineno">  872</span>&#160;{</div><div class="line"><a name="l00873"></a><span class="lineno">  873</span>&#160;    <a class="code" href="classarmnn_1_1_range_tracker.html">RangeTracker</a> ranges;</div><div class="line"><a name="l00874"></a><span class="lineno">  874</span>&#160;    <a class="code" href="classarmnn_1_1_range_tracker.html#a997e96288bdb106c922202e3f33d5d7b">RangeTracker::MinMaxRange</a> minMaxRange(-12.3f, 45.6f); <span class="comment">// Range to use for the override</span></div><div class="line"><a name="l00875"></a><span class="lineno">  875</span>&#160;</div><div class="line"><a name="l00876"></a><span class="lineno">  876</span>&#160;    <a class="code" href="classarmnn_1_1_network.html">Network</a> network; <span class="comment">// Empty network</span></div><div class="line"><a name="l00877"></a><span class="lineno">  877</span>&#160;    <span class="keyword">auto</span> inputLayers = network.<a class="code" href="classarmnn_1_1_network.html#afe0a4f719f9752a405e71878da7012ba">GetGraph</a>().<a class="code" href="classarmnn_1_1_graph.html#a919fb58873ef3a6549e4490e226f2eae">GetInputLayers</a>(); <span class="comment">// Empty list of input layers</span></div><div class="line"><a name="l00878"></a><span class="lineno">  878</span>&#160;</div><div class="line"><a name="l00879"></a><span class="lineno">  879</span>&#160;    <a class="code" href="classarmnn_1_1_override_input_range_visitor.html">OverrideInputRangeVisitor</a> overrideInputRangeVisitor(ranges, 0, minMaxRange);</div><div class="line"><a name="l00880"></a><span class="lineno">  880</span>&#160;    <a class="code" href="namespacearmnn.html#a9835ef753dda5b5a2fe827680e41fda7">VisitLayers</a>(inputLayers, overrideInputRangeVisitor);</div><div class="line"><a name="l00881"></a><span class="lineno">  881</span>&#160;</div><div class="line"><a name="l00882"></a><span class="lineno">  882</span>&#160;    <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a0c262ba6f6c189a2d092d127c1b7627b">BOOST_CHECK</a>(ranges.<a class="code" href="classarmnn_1_1_range_tracker.html#a8e12342fc420701fbffd97025421575a">IsEmpty</a>()); <span class="comment">// Check that the map of ranges remained untouched</span></div><div class="line"><a name="l00883"></a><span class="lineno">  883</span>&#160;}</div><div class="line"><a name="l00884"></a><span class="lineno">  884</span>&#160;</div><div class="line"><a name="l00885"></a><span class="lineno"><a class="line" href="namespacearmnn.html#ad432424d97021ae6c81e38130b1ec5d6">  885</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(OverrideInputRangeNoInputLayers)</div><div class="line"><a name="l00886"></a><span class="lineno">  886</span>&#160;{</div><div class="line"><a name="l00887"></a><span class="lineno">  887</span>&#160;    <a class="code" href="classarmnn_1_1_range_tracker.html">RangeTracker</a> ranges;</div><div class="line"><a name="l00888"></a><span class="lineno">  888</span>&#160;    <a class="code" href="namespacearmnn.html#a997e96288bdb106c922202e3f33d5d7b">MinMaxRange</a> minMaxRange(-12.3f, 45.6f); <span class="comment">// Range to use for the override</span></div><div class="line"><a name="l00889"></a><span class="lineno">  889</span>&#160;</div><div class="line"><a name="l00890"></a><span class="lineno">  890</span>&#160;    <a class="code" href="classarmnn_1_1_network.html">Network</a> network;</div><div class="line"><a name="l00891"></a><span class="lineno">  891</span>&#160;    network.<a class="code" href="classarmnn_1_1_network.html#adb60c75544796e23d7abc1ce0476f6d9">AddAdditionLayer</a>(); <span class="comment">// Network with no input layers</span></div><div class="line"><a name="l00892"></a><span class="lineno">  892</span>&#160;    <span class="keyword">auto</span> inputLayers = network.<a class="code" href="classarmnn_1_1_network.html#afe0a4f719f9752a405e71878da7012ba">GetGraph</a>().<a class="code" href="classarmnn_1_1_graph.html#a919fb58873ef3a6549e4490e226f2eae">GetInputLayers</a>(); <span class="comment">// Empty list of input layers</span></div><div class="line"><a name="l00893"></a><span class="lineno">  893</span>&#160;</div><div class="line"><a name="l00894"></a><span class="lineno">  894</span>&#160;    <a class="code" href="classarmnn_1_1_override_input_range_visitor.html">OverrideInputRangeVisitor</a> overrideInputRangeVisitor(ranges, 0, minMaxRange);</div><div class="line"><a name="l00895"></a><span class="lineno">  895</span>&#160;    <a class="code" href="namespacearmnn.html#a9835ef753dda5b5a2fe827680e41fda7">VisitLayers</a>(inputLayers, overrideInputRangeVisitor);</div><div class="line"><a name="l00896"></a><span class="lineno">  896</span>&#160;</div><div class="line"><a name="l00897"></a><span class="lineno">  897</span>&#160;    <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a0c262ba6f6c189a2d092d127c1b7627b">BOOST_CHECK</a>(ranges.<a class="code" href="classarmnn_1_1_range_tracker.html#a8e12342fc420701fbffd97025421575a">IsEmpty</a>()); <span class="comment">// Check that the map of ranges remained untouched</span></div><div class="line"><a name="l00898"></a><span class="lineno">  898</span>&#160;}</div><div class="line"><a name="l00899"></a><span class="lineno">  899</span>&#160;</div><div class="line"><a name="l00900"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a6e97e093453fc053a5c82386927a0d6c">  900</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(OverrideInputRangeInputLayers)</div><div class="line"><a name="l00901"></a><span class="lineno">  901</span>&#160;{</div><div class="line"><a name="l00902"></a><span class="lineno">  902</span>&#160;    <a class="code" href="classarmnn_1_1_range_tracker.html">RangeTracker</a> ranges;</div><div class="line"><a name="l00903"></a><span class="lineno">  903</span>&#160;    <a class="code" href="namespacearmnn.html#a997e96288bdb106c922202e3f33d5d7b">MinMaxRange</a> minMaxRange(-12.3f, 45.6f); <span class="comment">// Range to use for the override</span></div><div class="line"><a name="l00904"></a><span class="lineno">  904</span>&#160;</div><div class="line"><a name="l00905"></a><span class="lineno">  905</span>&#160;    <a class="code" href="classarmnn_1_1_network.html">Network</a> network;</div><div class="line"><a name="l00906"></a><span class="lineno">  906</span>&#160;</div><div class="line"><a name="l00907"></a><span class="lineno">  907</span>&#160;    <span class="comment">// Adding the layers</span></div><div class="line"><a name="l00908"></a><span class="lineno">  908</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input0 = network.<a class="code" href="classarmnn_1_1_network.html#a90d8841cfbbc82ab02328f33fed24ac6">AddInputLayer</a>(0);</div><div class="line"><a name="l00909"></a><span class="lineno">  909</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input1 = network.<a class="code" href="classarmnn_1_1_network.html#a90d8841cfbbc82ab02328f33fed24ac6">AddInputLayer</a>(1);</div><div class="line"><a name="l00910"></a><span class="lineno">  910</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* addition = network.<a class="code" href="classarmnn_1_1_network.html#adb60c75544796e23d7abc1ce0476f6d9">AddAdditionLayer</a>();</div><div class="line"><a name="l00911"></a><span class="lineno">  911</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network.<a class="code" href="classarmnn_1_1_network.html#ad55ff20f4c7e60c18b849e61f28f0e2e">AddOutputLayer</a>(2);</div><div class="line"><a name="l00912"></a><span class="lineno">  912</span>&#160;</div><div class="line"><a name="l00913"></a><span class="lineno">  913</span>&#160;    <span class="comment">// Connecting the layer</span></div><div class="line"><a name="l00914"></a><span class="lineno">  914</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l00915"></a><span class="lineno">  915</span>&#160;    input1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(1));</div><div class="line"><a name="l00916"></a><span class="lineno">  916</span>&#160;    addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l00917"></a><span class="lineno">  917</span>&#160;</div><div class="line"><a name="l00918"></a><span class="lineno">  918</span>&#160;    <span class="comment">// Setting the TensorInfos</span></div><div class="line"><a name="l00919"></a><span class="lineno">  919</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l00920"></a><span class="lineno">  920</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l00921"></a><span class="lineno">  921</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l00922"></a><span class="lineno">  922</span>&#160;    input1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l00923"></a><span class="lineno">  923</span>&#160;    addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l00924"></a><span class="lineno">  924</span>&#160;</div><div class="line"><a name="l00925"></a><span class="lineno">  925</span>&#160;    <span class="keyword">auto</span> inputLayers = network.<a class="code" href="classarmnn_1_1_network.html#afe0a4f719f9752a405e71878da7012ba">GetGraph</a>().<a class="code" href="classarmnn_1_1_graph.html#a919fb58873ef3a6549e4490e226f2eae">GetInputLayers</a>(); <span class="comment">// List of input layers</span></div><div class="line"><a name="l00926"></a><span class="lineno">  926</span>&#160;</div><div class="line"><a name="l00927"></a><span class="lineno">  927</span>&#160;    <span class="comment">// Trying to override the input range for the input layer with binding id 3 (does not exist in the network)</span></div><div class="line"><a name="l00928"></a><span class="lineno">  928</span>&#160;    <a class="code" href="classarmnn_1_1_override_input_range_visitor.html">OverrideInputRangeVisitor</a> overrideInputRangeVisitorLayer3(ranges, 3, minMaxRange);</div><div class="line"><a name="l00929"></a><span class="lineno">  929</span>&#160;    <a class="code" href="namespacearmnn.html#a9835ef753dda5b5a2fe827680e41fda7">VisitLayers</a>(inputLayers, overrideInputRangeVisitorLayer3);</div><div class="line"><a name="l00930"></a><span class="lineno">  930</span>&#160;</div><div class="line"><a name="l00931"></a><span class="lineno">  931</span>&#160;    <span class="comment">// Check that the map of ranges remained untouched</span></div><div class="line"><a name="l00932"></a><span class="lineno">  932</span>&#160;    <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a0c262ba6f6c189a2d092d127c1b7627b">BOOST_CHECK</a>(ranges.<a class="code" href="classarmnn_1_1_range_tracker.html#a8e12342fc420701fbffd97025421575a">IsEmpty</a>());</div><div class="line"><a name="l00933"></a><span class="lineno">  933</span>&#160;</div><div class="line"><a name="l00934"></a><span class="lineno">  934</span>&#160;    <span class="comment">// Override the input range for the input layer with binding id 1</span></div><div class="line"><a name="l00935"></a><span class="lineno">  935</span>&#160;    <a class="code" href="classarmnn_1_1_override_input_range_visitor.html">OverrideInputRangeVisitor</a> overrideInputRangeVisitorLayer1(ranges, 1, minMaxRange);</div><div class="line"><a name="l00936"></a><span class="lineno">  936</span>&#160;    <a class="code" href="namespacearmnn.html#a9835ef753dda5b5a2fe827680e41fda7">VisitLayers</a>(inputLayers, overrideInputRangeVisitorLayer1);</div><div class="line"><a name="l00937"></a><span class="lineno">  937</span>&#160;</div><div class="line"><a name="l00938"></a><span class="lineno">  938</span>&#160;    <span class="comment">// Check that the map of ranges has been populated</span></div><div class="line"><a name="l00939"></a><span class="lineno">  939</span>&#160;    <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a0c262ba6f6c189a2d092d127c1b7627b">BOOST_CHECK</a>(!ranges.<a class="code" href="classarmnn_1_1_range_tracker.html#a8e12342fc420701fbffd97025421575a">IsEmpty</a>());</div><div class="line"><a name="l00940"></a><span class="lineno">  940</span>&#160;</div><div class="line"><a name="l00941"></a><span class="lineno">  941</span>&#160;    <span class="comment">// Check that an entry for the input layer with binding id 0 does not exist</span></div><div class="line"><a name="l00942"></a><span class="lineno">  942</span>&#160;    <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a0c262ba6f6c189a2d092d127c1b7627b">BOOST_CHECK</a>(!ranges.<a class="code" href="classarmnn_1_1_range_tracker.html#a084c5aacd7e3bb07f2cfd5a8e9b0dd30">HasRanges</a>(input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#afb5e65c770f6cee222db8af7581541a6">GetGuid</a>()));</div><div class="line"><a name="l00943"></a><span class="lineno">  943</span>&#160;</div><div class="line"><a name="l00944"></a><span class="lineno">  944</span>&#160;    <span class="comment">// Check that an entry for the input layer with binding id 1 exists</span></div><div class="line"><a name="l00945"></a><span class="lineno">  945</span>&#160;    <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a0c262ba6f6c189a2d092d127c1b7627b">BOOST_CHECK</a>(ranges.<a class="code" href="classarmnn_1_1_range_tracker.html#a084c5aacd7e3bb07f2cfd5a8e9b0dd30">HasRanges</a>(input1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#afb5e65c770f6cee222db8af7581541a6">GetGuid</a>()));</div><div class="line"><a name="l00946"></a><span class="lineno">  946</span>&#160;</div><div class="line"><a name="l00947"></a><span class="lineno">  947</span>&#160;    <span class="comment">// Check the the overridden values are what we intended to set</span></div><div class="line"><a name="l00948"></a><span class="lineno">  948</span>&#160;    <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a0c262ba6f6c189a2d092d127c1b7627b">BOOST_CHECK</a>(ranges.<a class="code" href="classarmnn_1_1_range_tracker.html#a507bae23f59e94b4161886ebe663cdf4">GetRange</a>(input1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#afb5e65c770f6cee222db8af7581541a6">GetGuid</a>(), 0) == minMaxRange);</div><div class="line"><a name="l00949"></a><span class="lineno">  949</span>&#160;}</div><div class="line"><a name="l00950"></a><span class="lineno">  950</span>&#160;</div><div class="line"><a name="l00951"></a><span class="lineno"><a class="line" href="namespacearmnn.html#aad4b8cb9a4d882a48bc21510f0d1a938">  951</a></span>&#160;<a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> <a class="code" href="namespacearmnn.html#aad4b8cb9a4d882a48bc21510f0d1a938">CreateNetworkWithFullyConnectedLayer</a>(<span class="keyword">const</span> <span class="keywordtype">bool</span> biasEnabled,</div><div class="line"><a name="l00952"></a><span class="lineno">  952</span>&#160;                                                 <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l00953"></a><span class="lineno">  953</span>&#160;                                                 <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00954"></a><span class="lineno">  954</span>&#160;{</div><div class="line"><a name="l00955"></a><span class="lineno">  955</span>&#160;    <a class="code" href="structarmnn_1_1_fully_connected_descriptor.html">FullyConnectedDescriptor</a> desc;</div><div class="line"><a name="l00956"></a><span class="lineno">  956</span>&#160;    desc.<a class="code" href="structarmnn_1_1_fully_connected_descriptor.html#aea202e14d8874cefd9a0f778022b7e25">m_BiasEnabled</a> = biasEnabled;</div><div class="line"><a name="l00957"></a><span class="lineno">  957</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l00958"></a><span class="lineno">  958</span>&#160;</div><div class="line"><a name="l00959"></a><span class="lineno">  959</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(inputShape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l00960"></a><span class="lineno">  960</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> outputInfo(outputShape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l00961"></a><span class="lineno">  961</span>&#160;</div><div class="line"><a name="l00962"></a><span class="lineno">  962</span>&#160;    std::vector&lt;float&gt; weightsData{-1.0f, 1.5f, 2.0f};</div><div class="line"><a name="l00963"></a><span class="lineno">  963</span>&#160;    <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a> weights(info, weightsData);</div><div class="line"><a name="l00964"></a><span class="lineno">  964</span>&#160;</div><div class="line"><a name="l00965"></a><span class="lineno">  965</span>&#160;    <span class="comment">// Add the layers</span></div><div class="line"><a name="l00966"></a><span class="lineno">  966</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input0 = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l00967"></a><span class="lineno">  967</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* fullyConnected;</div><div class="line"><a name="l00968"></a><span class="lineno">  968</span>&#160;    <a class="code" href="classarmnn_1_1_optional.html">Optional&lt;ConstTensor&gt;</a> optionalBias;</div><div class="line"><a name="l00969"></a><span class="lineno">  969</span>&#160;    std::vector&lt;float&gt; biasData{10.0f, 20.0f, 30.0f};</div><div class="line"><a name="l00970"></a><span class="lineno">  970</span>&#160;    <span class="keywordflow">if</span> (desc.<a class="code" href="structarmnn_1_1_fully_connected_descriptor.html#aea202e14d8874cefd9a0f778022b7e25">m_BiasEnabled</a>)</div><div class="line"><a name="l00971"></a><span class="lineno">  971</span>&#160;    {</div><div class="line"><a name="l00972"></a><span class="lineno">  972</span>&#160;        <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a> bias(info, biasData);</div><div class="line"><a name="l00973"></a><span class="lineno">  973</span>&#160;        optionalBias = <a class="code" href="classarmnn_1_1_optional.html">Optional&lt;ConstTensor&gt;</a>(bias);</div><div class="line"><a name="l00974"></a><span class="lineno">  974</span>&#160;    }</div><div class="line"><a name="l00975"></a><span class="lineno">  975</span>&#160;    fullyConnected = network-&gt;AddFullyConnectedLayer(desc, weights, optionalBias);</div><div class="line"><a name="l00976"></a><span class="lineno">  976</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;AddOutputLayer(1);</div><div class="line"><a name="l00977"></a><span class="lineno">  977</span>&#160;</div><div class="line"><a name="l00978"></a><span class="lineno">  978</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l00979"></a><span class="lineno">  979</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(fullyConnected-&gt;GetInputSlot(0));</div><div class="line"><a name="l00980"></a><span class="lineno">  980</span>&#160;    fullyConnected-&gt;GetOutputSlot(0).Connect(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l00981"></a><span class="lineno">  981</span>&#160;</div><div class="line"><a name="l00982"></a><span class="lineno">  982</span>&#160;    <span class="comment">// Set TensorInfo</span></div><div class="line"><a name="l00983"></a><span class="lineno">  983</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l00984"></a><span class="lineno">  984</span>&#160;    fullyConnected-&gt;GetOutputSlot(0).SetTensorInfo(outputInfo);</div><div class="line"><a name="l00985"></a><span class="lineno">  985</span>&#160;</div><div class="line"><a name="l00986"></a><span class="lineno">  986</span>&#160;    <span class="keywordflow">return</span> network;</div><div class="line"><a name="l00987"></a><span class="lineno">  987</span>&#160;}</div><div class="line"><a name="l00988"></a><span class="lineno">  988</span>&#160;</div><div class="line"><a name="l00989"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a245661fc96c9c4a9b898e1d98c8c6962">  989</a></span>&#160;<span class="keywordtype">void</span> <a class="code" href="namespacearmnn.html#a245661fc96c9c4a9b898e1d98c8c6962">ValidateFullyConnectedLayer</a>(<span class="keyword">const</span> <span class="keywordtype">bool</span> biasEnabled)</div><div class="line"><a name="l00990"></a><span class="lineno">  990</span>&#160;{</div><div class="line"><a name="l00991"></a><span class="lineno">  991</span>&#160;    <span class="keyword">class </span>TestFullyConnectedQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l00992"></a><span class="lineno">  992</span>&#160;    {</div><div class="line"><a name="l00993"></a><span class="lineno">  993</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l00994"></a><span class="lineno">  994</span>&#160;        TestFullyConnectedQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l00995"></a><span class="lineno">  995</span>&#160;        : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l00996"></a><span class="lineno">  996</span>&#160;</div><div class="line"><a name="l00997"></a><span class="lineno">  997</span>&#160;        TestFullyConnectedQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l00998"></a><span class="lineno">  998</span>&#160;                                       <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l00999"></a><span class="lineno">  999</span>&#160;                                       <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01000"></a><span class="lineno"> 1000</span>&#160;        : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l01001"></a><span class="lineno"> 1001</span>&#160;</div><div class="line"><a name="l01002"></a><span class="lineno"> 1002</span>&#160;        <span class="keywordtype">void</span> VisitFullyConnectedLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l01003"></a><span class="lineno"> 1003</span>&#160;                                      <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_fully_connected_descriptor.html">FullyConnectedDescriptor</a>&amp; desc,</div><div class="line"><a name="l01004"></a><span class="lineno"> 1004</span>&#160;                                      <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a>&amp; weights,</div><div class="line"><a name="l01005"></a><span class="lineno"> 1005</span>&#160;                                      <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_optional.html">Optional&lt;ConstTensor&gt;</a>&amp; biases,</div><div class="line"><a name="l01006"></a><span class="lineno"> 1006</span>&#160;                                      <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01007"></a><span class="lineno"> 1007</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01008"></a><span class="lineno"> 1008</span>&#160;            boost::ignore_unused(desc, name);</div><div class="line"><a name="l01009"></a><span class="lineno"> 1009</span>&#160;            TestQuantizationOnLayersWithBiases(layer, weights, biases);</div><div class="line"><a name="l01010"></a><span class="lineno"> 1010</span>&#160;        }</div><div class="line"><a name="l01011"></a><span class="lineno"> 1011</span>&#160;    };</div><div class="line"><a name="l01012"></a><span class="lineno"> 1012</span>&#160;</div><div class="line"><a name="l01013"></a><span class="lineno"> 1013</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{3U};</div><div class="line"><a name="l01014"></a><span class="lineno"> 1014</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="namespacearmnn.html#aad4b8cb9a4d882a48bc21510f0d1a938">CreateNetworkWithFullyConnectedLayer</a>(biasEnabled, shape, shape);</div><div class="line"><a name="l01015"></a><span class="lineno"> 1015</span>&#160;</div><div class="line"><a name="l01016"></a><span class="lineno"> 1016</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l01017"></a><span class="lineno"> 1017</span>&#160;    TestFullyConnectedQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l01018"></a><span class="lineno"> 1018</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l01019"></a><span class="lineno"> 1019</span>&#160;</div><div class="line"><a name="l01020"></a><span class="lineno"> 1020</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l01021"></a><span class="lineno"> 1021</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01022"></a><span class="lineno"> 1022</span>&#160;    TestFullyConnectedQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l01023"></a><span class="lineno"> 1023</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l01024"></a><span class="lineno"> 1024</span>&#160;</div><div class="line"><a name="l01025"></a><span class="lineno"> 1025</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l01026"></a><span class="lineno"> 1026</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01027"></a><span class="lineno"> 1027</span>&#160;    TestFullyConnectedQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l01028"></a><span class="lineno"> 1028</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l01029"></a><span class="lineno"> 1029</span>&#160;</div><div class="line"><a name="l01030"></a><span class="lineno"> 1030</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> Qsymm16Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l01031"></a><span class="lineno"> 1031</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), Qsymm16Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01032"></a><span class="lineno"> 1032</span>&#160;    TestFullyConnectedQuantization validatorQSymmS16(Qsymm16Options, shape, shape);</div><div class="line"><a name="l01033"></a><span class="lineno"> 1033</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l01034"></a><span class="lineno"> 1034</span>&#160;}</div><div class="line"><a name="l01035"></a><span class="lineno"> 1035</span>&#160;</div><div class="line"><a name="l01036"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a881ab05533f917737509402730668e4a"> 1036</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeFullyConnected)</div><div class="line"><a name="l01037"></a><span class="lineno"> 1037</span>&#160;{</div><div class="line"><a name="l01038"></a><span class="lineno"> 1038</span>&#160;    <a class="code" href="namespacearmnn.html#a245661fc96c9c4a9b898e1d98c8c6962">ValidateFullyConnectedLayer</a>(<span class="keyword">false</span>);</div><div class="line"><a name="l01039"></a><span class="lineno"> 1039</span>&#160;}</div><div class="line"><a name="l01040"></a><span class="lineno"> 1040</span>&#160;</div><div class="line"><a name="l01041"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a69dd8c7608ff0935a247f3aa07f98212"> 1041</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeFullyConnectedBiasEnabled)</div><div class="line"><a name="l01042"></a><span class="lineno"> 1042</span>&#160;{</div><div class="line"><a name="l01043"></a><span class="lineno"> 1043</span>&#160;    <a class="code" href="namespacearmnn.html#a245661fc96c9c4a9b898e1d98c8c6962">ValidateFullyConnectedLayer</a>(<span class="keyword">true</span>);</div><div class="line"><a name="l01044"></a><span class="lineno"> 1044</span>&#160;}</div><div class="line"><a name="l01045"></a><span class="lineno"> 1045</span>&#160;</div><div class="line"><a name="l01046"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a14cfd39cfc30682fa821ade3dd298426"> 1046</a></span>&#160;<span class="keywordtype">void</span> <a class="code" href="namespacearmnn.html#a14cfd39cfc30682fa821ade3dd298426">TestQuantizeConvolution2d</a>(<span class="keywordtype">bool</span> useBiases)</div><div class="line"><a name="l01047"></a><span class="lineno"> 1047</span>&#160;{</div><div class="line"><a name="l01048"></a><span class="lineno"> 1048</span>&#160;    <span class="keyword">class </span>TestConv2dQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l01049"></a><span class="lineno"> 1049</span>&#160;    {</div><div class="line"><a name="l01050"></a><span class="lineno"> 1050</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l01051"></a><span class="lineno"> 1051</span>&#160;        TestConv2dQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01052"></a><span class="lineno"> 1052</span>&#160;        : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l01053"></a><span class="lineno"> 1053</span>&#160;</div><div class="line"><a name="l01054"></a><span class="lineno"> 1054</span>&#160;        TestConv2dQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l01055"></a><span class="lineno"> 1055</span>&#160;                               <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l01056"></a><span class="lineno"> 1056</span>&#160;                               <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01057"></a><span class="lineno"> 1057</span>&#160;        : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l01058"></a><span class="lineno"> 1058</span>&#160;</div><div class="line"><a name="l01059"></a><span class="lineno"> 1059</span>&#160;        <span class="keywordtype">void</span> VisitConvolution2dLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a> *layer,</div><div class="line"><a name="l01060"></a><span class="lineno"> 1060</span>&#160;                                     <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_convolution2d_descriptor.html">Convolution2dDescriptor</a>&amp; convolution2dDescriptor,</div><div class="line"><a name="l01061"></a><span class="lineno"> 1061</span>&#160;                                     <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a>&amp; weights,</div><div class="line"><a name="l01062"></a><span class="lineno"> 1062</span>&#160;                                     <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_optional.html">Optional&lt;ConstTensor&gt;</a>&amp; biases,</div><div class="line"><a name="l01063"></a><span class="lineno"> 1063</span>&#160;                                     <span class="keyword">const</span> <span class="keywordtype">char</span> *name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01064"></a><span class="lineno"> 1064</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01065"></a><span class="lineno"> 1065</span>&#160;            boost::ignore_unused(convolution2dDescriptor, name);</div><div class="line"><a name="l01066"></a><span class="lineno"> 1066</span>&#160;            TestQuantizationOnLayersWithBiases(layer, weights, biases);</div><div class="line"><a name="l01067"></a><span class="lineno"> 1067</span>&#160;        }</div><div class="line"><a name="l01068"></a><span class="lineno"> 1068</span>&#160;    };</div><div class="line"><a name="l01069"></a><span class="lineno"> 1069</span>&#160;</div><div class="line"><a name="l01070"></a><span class="lineno"> 1070</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l01071"></a><span class="lineno"> 1071</span>&#160;</div><div class="line"><a name="l01072"></a><span class="lineno"> 1072</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{3U};</div><div class="line"><a name="l01073"></a><span class="lineno"> 1073</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l01074"></a><span class="lineno"> 1074</span>&#160;</div><div class="line"><a name="l01075"></a><span class="lineno"> 1075</span>&#160;    std::vector&lt;float&gt; weightsData{-1.0f, 1.5f, 2.0f};</div><div class="line"><a name="l01076"></a><span class="lineno"> 1076</span>&#160;    <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a> weights(info, weightsData);</div><div class="line"><a name="l01077"></a><span class="lineno"> 1077</span>&#160;</div><div class="line"><a name="l01078"></a><span class="lineno"> 1078</span>&#160;    <a class="code" href="structarmnn_1_1_convolution2d_descriptor.html">Convolution2dDescriptor</a> descriptor;</div><div class="line"><a name="l01079"></a><span class="lineno"> 1079</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_convolution2d_descriptor.html#aea202e14d8874cefd9a0f778022b7e25">m_BiasEnabled</a> = useBiases;</div><div class="line"><a name="l01080"></a><span class="lineno"> 1080</span>&#160;</div><div class="line"><a name="l01081"></a><span class="lineno"> 1081</span>&#160;    <span class="comment">// Add the layers</span></div><div class="line"><a name="l01082"></a><span class="lineno"> 1082</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input0 = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l01083"></a><span class="lineno"> 1083</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* conv2d;</div><div class="line"><a name="l01084"></a><span class="lineno"> 1084</span>&#160;    <a class="code" href="classarmnn_1_1_optional.html">Optional&lt;ConstTensor&gt;</a> optionalBiases;</div><div class="line"><a name="l01085"></a><span class="lineno"> 1085</span>&#160;    std::vector&lt;float&gt; biasesData{-1.0f, 1.5f, 2.0f};</div><div class="line"><a name="l01086"></a><span class="lineno"> 1086</span>&#160;    <span class="keywordflow">if</span> (useBiases)</div><div class="line"><a name="l01087"></a><span class="lineno"> 1087</span>&#160;    {</div><div class="line"><a name="l01088"></a><span class="lineno"> 1088</span>&#160;        <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a> biases(info, biasesData);</div><div class="line"><a name="l01089"></a><span class="lineno"> 1089</span>&#160;        optionalBiases = <a class="code" href="classarmnn_1_1_optional.html">Optional&lt;ConstTensor&gt;</a>(biases);</div><div class="line"><a name="l01090"></a><span class="lineno"> 1090</span>&#160;    }</div><div class="line"><a name="l01091"></a><span class="lineno"> 1091</span>&#160;    conv2d = network-&gt;AddConvolution2dLayer(descriptor, weights, optionalBiases);</div><div class="line"><a name="l01092"></a><span class="lineno"> 1092</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;AddOutputLayer(1);</div><div class="line"><a name="l01093"></a><span class="lineno"> 1093</span>&#160;</div><div class="line"><a name="l01094"></a><span class="lineno"> 1094</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l01095"></a><span class="lineno"> 1095</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(conv2d-&gt;GetInputSlot(0));</div><div class="line"><a name="l01096"></a><span class="lineno"> 1096</span>&#160;    conv2d-&gt;GetOutputSlot(0).Connect(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01097"></a><span class="lineno"> 1097</span>&#160;</div><div class="line"><a name="l01098"></a><span class="lineno"> 1098</span>&#160;    <span class="comment">// Set TensorInfo</span></div><div class="line"><a name="l01099"></a><span class="lineno"> 1099</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l01100"></a><span class="lineno"> 1100</span>&#160;    conv2d-&gt;GetOutputSlot(0).SetTensorInfo(info);</div><div class="line"><a name="l01101"></a><span class="lineno"> 1101</span>&#160;</div><div class="line"><a name="l01102"></a><span class="lineno"> 1102</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l01103"></a><span class="lineno"> 1103</span>&#160;    TestConv2dQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l01104"></a><span class="lineno"> 1104</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l01105"></a><span class="lineno"> 1105</span>&#160;</div><div class="line"><a name="l01106"></a><span class="lineno"> 1106</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l01107"></a><span class="lineno"> 1107</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01108"></a><span class="lineno"> 1108</span>&#160;    TestConv2dQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l01109"></a><span class="lineno"> 1109</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l01110"></a><span class="lineno"> 1110</span>&#160;</div><div class="line"><a name="l01111"></a><span class="lineno"> 1111</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l01112"></a><span class="lineno"> 1112</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01113"></a><span class="lineno"> 1113</span>&#160;    TestConv2dQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l01114"></a><span class="lineno"> 1114</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l01115"></a><span class="lineno"> 1115</span>&#160;</div><div class="line"><a name="l01116"></a><span class="lineno"> 1116</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> Qsymm16Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l01117"></a><span class="lineno"> 1117</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), Qsymm16Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01118"></a><span class="lineno"> 1118</span>&#160;    TestConv2dQuantization validatorQSymmS16(Qsymm16Options, shape, shape);</div><div class="line"><a name="l01119"></a><span class="lineno"> 1119</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l01120"></a><span class="lineno"> 1120</span>&#160;}</div><div class="line"><a name="l01121"></a><span class="lineno"> 1121</span>&#160;</div><div class="line"><a name="l01122"></a><span class="lineno"><a class="line" href="namespacearmnn.html#aa117e0112fdc02a7a011cfb39dc596ab"> 1122</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeConvolution2d)</div><div class="line"><a name="l01123"></a><span class="lineno"> 1123</span>&#160;{</div><div class="line"><a name="l01124"></a><span class="lineno"> 1124</span>&#160;    <a class="code" href="namespacearmnn.html#a14cfd39cfc30682fa821ade3dd298426">TestQuantizeConvolution2d</a>(<span class="keyword">false</span>);</div><div class="line"><a name="l01125"></a><span class="lineno"> 1125</span>&#160;}</div><div class="line"><a name="l01126"></a><span class="lineno"> 1126</span>&#160;</div><div class="line"><a name="l01127"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a9827adb2cf787460578999e0484568fa"> 1127</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeConvolution2dWithBiases)</div><div class="line"><a name="l01128"></a><span class="lineno"> 1128</span>&#160;{</div><div class="line"><a name="l01129"></a><span class="lineno"> 1129</span>&#160;    <a class="code" href="namespacearmnn.html#a14cfd39cfc30682fa821ade3dd298426">TestQuantizeConvolution2d</a>(<span class="keyword">true</span>);</div><div class="line"><a name="l01130"></a><span class="lineno"> 1130</span>&#160;}</div><div class="line"><a name="l01131"></a><span class="lineno"> 1131</span>&#160;</div><div class="line"><a name="l01132"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a5abbe8a9ee003c1379a921dbe2745b81"> 1132</a></span>&#160;<span class="keywordtype">void</span> <a class="code" href="namespacearmnn.html#a5abbe8a9ee003c1379a921dbe2745b81">TestQuantizeDepthwiseConvolution2d</a>(<span class="keywordtype">bool</span> useBiases)</div><div class="line"><a name="l01133"></a><span class="lineno"> 1133</span>&#160;{</div><div class="line"><a name="l01134"></a><span class="lineno"> 1134</span>&#160;    <span class="keyword">class </span>TestDepthwiseConv2dQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l01135"></a><span class="lineno"> 1135</span>&#160;    {</div><div class="line"><a name="l01136"></a><span class="lineno"> 1136</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l01137"></a><span class="lineno"> 1137</span>&#160;        TestDepthwiseConv2dQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01138"></a><span class="lineno"> 1138</span>&#160;        : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l01139"></a><span class="lineno"> 1139</span>&#160;</div><div class="line"><a name="l01140"></a><span class="lineno"> 1140</span>&#160;        TestDepthwiseConv2dQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l01141"></a><span class="lineno"> 1141</span>&#160;                                        <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l01142"></a><span class="lineno"> 1142</span>&#160;                                        <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01143"></a><span class="lineno"> 1143</span>&#160;        : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l01144"></a><span class="lineno"> 1144</span>&#160;</div><div class="line"><a name="l01145"></a><span class="lineno"> 1145</span>&#160;        <span class="keywordtype">void</span> VisitDepthwiseConvolution2dLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a> *layer,</div><div class="line"><a name="l01146"></a><span class="lineno"> 1146</span>&#160;                                              <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_depthwise_convolution2d_descriptor.html">DepthwiseConvolution2dDescriptor</a>&amp; convolution2dDescriptor,</div><div class="line"><a name="l01147"></a><span class="lineno"> 1147</span>&#160;                                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a>&amp; weights,</div><div class="line"><a name="l01148"></a><span class="lineno"> 1148</span>&#160;                                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_optional.html">Optional&lt;ConstTensor&gt;</a>&amp; biases,</div><div class="line"><a name="l01149"></a><span class="lineno"> 1149</span>&#160;                                              <span class="keyword">const</span> <span class="keywordtype">char</span> *name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01150"></a><span class="lineno"> 1150</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01151"></a><span class="lineno"> 1151</span>&#160;            boost::ignore_unused(convolution2dDescriptor, name);</div><div class="line"><a name="l01152"></a><span class="lineno"> 1152</span>&#160;            TestQuantizationOnLayersWithBiases(layer, weights, biases);</div><div class="line"><a name="l01153"></a><span class="lineno"> 1153</span>&#160;        }</div><div class="line"><a name="l01154"></a><span class="lineno"> 1154</span>&#160;    };</div><div class="line"><a name="l01155"></a><span class="lineno"> 1155</span>&#160;</div><div class="line"><a name="l01156"></a><span class="lineno"> 1156</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l01157"></a><span class="lineno"> 1157</span>&#160;</div><div class="line"><a name="l01158"></a><span class="lineno"> 1158</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{3U};</div><div class="line"><a name="l01159"></a><span class="lineno"> 1159</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l01160"></a><span class="lineno"> 1160</span>&#160;</div><div class="line"><a name="l01161"></a><span class="lineno"> 1161</span>&#160;    std::vector&lt;float&gt; weightsData{-1.0f, 1.5f, 2.0f};</div><div class="line"><a name="l01162"></a><span class="lineno"> 1162</span>&#160;    <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a> weights(info, weightsData);</div><div class="line"><a name="l01163"></a><span class="lineno"> 1163</span>&#160;</div><div class="line"><a name="l01164"></a><span class="lineno"> 1164</span>&#160;    <a class="code" href="structarmnn_1_1_depthwise_convolution2d_descriptor.html">DepthwiseConvolution2dDescriptor</a> descriptor;</div><div class="line"><a name="l01165"></a><span class="lineno"> 1165</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_depthwise_convolution2d_descriptor.html#aea202e14d8874cefd9a0f778022b7e25">m_BiasEnabled</a> = useBiases;</div><div class="line"><a name="l01166"></a><span class="lineno"> 1166</span>&#160;</div><div class="line"><a name="l01167"></a><span class="lineno"> 1167</span>&#160;    <span class="comment">// Add the layers</span></div><div class="line"><a name="l01168"></a><span class="lineno"> 1168</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input0 = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l01169"></a><span class="lineno"> 1169</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* depthwiseConv2d;</div><div class="line"><a name="l01170"></a><span class="lineno"> 1170</span>&#160;    <a class="code" href="classarmnn_1_1_optional.html">Optional&lt;ConstTensor&gt;</a> optionalBiases;</div><div class="line"><a name="l01171"></a><span class="lineno"> 1171</span>&#160;    std::vector&lt;float&gt; biasesData{-1.0f, 1.5f, 2.0f};</div><div class="line"><a name="l01172"></a><span class="lineno"> 1172</span>&#160;    <span class="keywordflow">if</span> (useBiases)</div><div class="line"><a name="l01173"></a><span class="lineno"> 1173</span>&#160;    {</div><div class="line"><a name="l01174"></a><span class="lineno"> 1174</span>&#160;        <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a> biases(info, biasesData);</div><div class="line"><a name="l01175"></a><span class="lineno"> 1175</span>&#160;        optionalBiases = <a class="code" href="classarmnn_1_1_optional.html">Optional&lt;ConstTensor&gt;</a>(biases);</div><div class="line"><a name="l01176"></a><span class="lineno"> 1176</span>&#160;    }</div><div class="line"><a name="l01177"></a><span class="lineno"> 1177</span>&#160;    depthwiseConv2d = network-&gt;AddDepthwiseConvolution2dLayer(descriptor, weights, optionalBiases);</div><div class="line"><a name="l01178"></a><span class="lineno"> 1178</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;AddOutputLayer(1);</div><div class="line"><a name="l01179"></a><span class="lineno"> 1179</span>&#160;</div><div class="line"><a name="l01180"></a><span class="lineno"> 1180</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l01181"></a><span class="lineno"> 1181</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(depthwiseConv2d-&gt;GetInputSlot(0));</div><div class="line"><a name="l01182"></a><span class="lineno"> 1182</span>&#160;    depthwiseConv2d-&gt;GetOutputSlot(0).Connect(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01183"></a><span class="lineno"> 1183</span>&#160;</div><div class="line"><a name="l01184"></a><span class="lineno"> 1184</span>&#160;    <span class="comment">//Set TensorInfo</span></div><div class="line"><a name="l01185"></a><span class="lineno"> 1185</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l01186"></a><span class="lineno"> 1186</span>&#160;    depthwiseConv2d-&gt;GetOutputSlot(0).SetTensorInfo(info);</div><div class="line"><a name="l01187"></a><span class="lineno"> 1187</span>&#160;</div><div class="line"><a name="l01188"></a><span class="lineno"> 1188</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l01189"></a><span class="lineno"> 1189</span>&#160;    TestDepthwiseConv2dQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l01190"></a><span class="lineno"> 1190</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l01191"></a><span class="lineno"> 1191</span>&#160;</div><div class="line"><a name="l01192"></a><span class="lineno"> 1192</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l01193"></a><span class="lineno"> 1193</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01194"></a><span class="lineno"> 1194</span>&#160;    TestDepthwiseConv2dQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l01195"></a><span class="lineno"> 1195</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l01196"></a><span class="lineno"> 1196</span>&#160;</div><div class="line"><a name="l01197"></a><span class="lineno"> 1197</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l01198"></a><span class="lineno"> 1198</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01199"></a><span class="lineno"> 1199</span>&#160;    TestDepthwiseConv2dQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l01200"></a><span class="lineno"> 1200</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l01201"></a><span class="lineno"> 1201</span>&#160;</div><div class="line"><a name="l01202"></a><span class="lineno"> 1202</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> Qsymm16Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l01203"></a><span class="lineno"> 1203</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), Qsymm16Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01204"></a><span class="lineno"> 1204</span>&#160;    TestDepthwiseConv2dQuantization validatorQSymmS16(Qsymm16Options, shape, shape);</div><div class="line"><a name="l01205"></a><span class="lineno"> 1205</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l01206"></a><span class="lineno"> 1206</span>&#160;}</div><div class="line"><a name="l01207"></a><span class="lineno"> 1207</span>&#160;</div><div class="line"><a name="l01208"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a1db5d836b83fc71602a7993616de5b42"> 1208</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeDepthwiseConvolution2d)</div><div class="line"><a name="l01209"></a><span class="lineno"> 1209</span>&#160;{</div><div class="line"><a name="l01210"></a><span class="lineno"> 1210</span>&#160;    <a class="code" href="namespacearmnn.html#a5abbe8a9ee003c1379a921dbe2745b81">TestQuantizeDepthwiseConvolution2d</a>(<span class="keyword">false</span>);</div><div class="line"><a name="l01211"></a><span class="lineno"> 1211</span>&#160;}</div><div class="line"><a name="l01212"></a><span class="lineno"> 1212</span>&#160;</div><div class="line"><a name="l01213"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a891abdb9079715cbcf85792e2b450652"> 1213</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeDepthwiseConvolution2dWithBiases)</div><div class="line"><a name="l01214"></a><span class="lineno"> 1214</span>&#160;{</div><div class="line"><a name="l01215"></a><span class="lineno"> 1215</span>&#160;    <a class="code" href="namespacearmnn.html#a5abbe8a9ee003c1379a921dbe2745b81">TestQuantizeDepthwiseConvolution2d</a>(<span class="keyword">true</span>);</div><div class="line"><a name="l01216"></a><span class="lineno"> 1216</span>&#160;}</div><div class="line"><a name="l01217"></a><span class="lineno"> 1217</span>&#160;</div><div class="line"><a name="l01218"></a><span class="lineno"><a class="line" href="namespacearmnn.html#abd033569519fec65077ea983f6c75a9d"> 1218</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeInstanceNormalization)</div><div class="line"><a name="l01219"></a><span class="lineno"> 1219</span>&#160;{</div><div class="line"><a name="l01220"></a><span class="lineno"> 1220</span>&#160;    <span class="keyword">class </span>TestInstanceNormalizationQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l01221"></a><span class="lineno"> 1221</span>&#160;    {</div><div class="line"><a name="l01222"></a><span class="lineno"> 1222</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l01223"></a><span class="lineno"> 1223</span>&#160;        TestInstanceNormalizationQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01224"></a><span class="lineno"> 1224</span>&#160;            : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l01225"></a><span class="lineno"> 1225</span>&#160;</div><div class="line"><a name="l01226"></a><span class="lineno"> 1226</span>&#160;        TestInstanceNormalizationQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l01227"></a><span class="lineno"> 1227</span>&#160;                                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l01228"></a><span class="lineno"> 1228</span>&#160;                                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01229"></a><span class="lineno"> 1229</span>&#160;            : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l01230"></a><span class="lineno"> 1230</span>&#160;</div><div class="line"><a name="l01231"></a><span class="lineno"> 1231</span>&#160;        <span class="keyword">virtual</span> <span class="keywordtype">void</span> VisitInstanceNormalizationLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l01232"></a><span class="lineno"> 1232</span>&#160;                                                     <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_instance_normalization_descriptor.html">InstanceNormalizationDescriptor</a>&amp; descriptor,</div><div class="line"><a name="l01233"></a><span class="lineno"> 1233</span>&#160;                                                     <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)</div><div class="line"><a name="l01234"></a><span class="lineno"> 1234</span>&#160;        {</div><div class="line"><a name="l01235"></a><span class="lineno"> 1235</span>&#160;            boost::ignore_unused(descriptor, name);</div><div class="line"><a name="l01236"></a><span class="lineno"> 1236</span>&#160;            <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l01237"></a><span class="lineno"> 1237</span>&#160;</div><div class="line"><a name="l01238"></a><span class="lineno"> 1238</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qAsymmU8Params{ 30.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 128 };</div><div class="line"><a name="l01239"></a><span class="lineno"> 1239</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qAsymmS8Params { 30.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>,  0};</div><div class="line"><a name="l01240"></a><span class="lineno"> 1240</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qSymmS8Params { 15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0};</div><div class="line"><a name="l01241"></a><span class="lineno"> 1241</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qSymmS16Params{ 15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0 };</div><div class="line"><a name="l01242"></a><span class="lineno"> 1242</span>&#160;</div><div class="line"><a name="l01243"></a><span class="lineno"> 1243</span>&#160;            TestQuantizationParams(info, qAsymmU8Params, qAsymmS8Params, qSymmS8Params, qSymmS16Params);</div><div class="line"><a name="l01244"></a><span class="lineno"> 1244</span>&#160;        }</div><div class="line"><a name="l01245"></a><span class="lineno"> 1245</span>&#160;    };</div><div class="line"><a name="l01246"></a><span class="lineno"> 1246</span>&#160;</div><div class="line"><a name="l01247"></a><span class="lineno"> 1247</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> tensorShape{ 1, 4, 4, 1 };</div><div class="line"><a name="l01248"></a><span class="lineno"> 1248</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> tensorInfo(tensorShape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l01249"></a><span class="lineno"> 1249</span>&#160;</div><div class="line"><a name="l01250"></a><span class="lineno"> 1250</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l01251"></a><span class="lineno"> 1251</span>&#160;</div><div class="line"><a name="l01252"></a><span class="lineno"> 1252</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* inputLayer        = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l01253"></a><span class="lineno"> 1253</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* instanceNormLayer = network-&gt;AddInstanceNormalizationLayer(<a class="code" href="structarmnn_1_1_instance_normalization_descriptor.html">InstanceNormalizationDescriptor</a>());</div><div class="line"><a name="l01254"></a><span class="lineno"> 1254</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* outputLayer       = network-&gt;AddOutputLayer(0);</div><div class="line"><a name="l01255"></a><span class="lineno"> 1255</span>&#160;</div><div class="line"><a name="l01256"></a><span class="lineno"> 1256</span>&#160;    inputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(instanceNormLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01257"></a><span class="lineno"> 1257</span>&#160;    instanceNormLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(outputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01258"></a><span class="lineno"> 1258</span>&#160;</div><div class="line"><a name="l01259"></a><span class="lineno"> 1259</span>&#160;    inputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(tensorInfo);</div><div class="line"><a name="l01260"></a><span class="lineno"> 1260</span>&#160;    instanceNormLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(tensorInfo);</div><div class="line"><a name="l01261"></a><span class="lineno"> 1261</span>&#160;</div><div class="line"><a name="l01262"></a><span class="lineno"> 1262</span>&#160;    <span class="comment">// test QAsymmU8 quantization</span></div><div class="line"><a name="l01263"></a><span class="lineno"> 1263</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l01264"></a><span class="lineno"> 1264</span>&#160;    TestInstanceNormalizationQuantization validatorQAsymmU8(tensorShape, tensorShape);</div><div class="line"><a name="l01265"></a><span class="lineno"> 1265</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l01266"></a><span class="lineno"> 1266</span>&#160;</div><div class="line"><a name="l01267"></a><span class="lineno"> 1267</span>&#160;    <span class="comment">//test QAsymmS8 quantization</span></div><div class="line"><a name="l01268"></a><span class="lineno"> 1268</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l01269"></a><span class="lineno"> 1269</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01270"></a><span class="lineno"> 1270</span>&#160;    TestInstanceNormalizationQuantization validatorQAsymmS8(qAsymmS8Options, tensorShape, tensorShape);</div><div class="line"><a name="l01271"></a><span class="lineno"> 1271</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l01272"></a><span class="lineno"> 1272</span>&#160;</div><div class="line"><a name="l01273"></a><span class="lineno"> 1273</span>&#160;    <span class="comment">// test QSymmS8 quantization</span></div><div class="line"><a name="l01274"></a><span class="lineno"> 1274</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l01275"></a><span class="lineno"> 1275</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01276"></a><span class="lineno"> 1276</span>&#160;    TestInstanceNormalizationQuantization validatorQSymmS8(qSymmS8Options, tensorShape, tensorShape);</div><div class="line"><a name="l01277"></a><span class="lineno"> 1277</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l01278"></a><span class="lineno"> 1278</span>&#160;</div><div class="line"><a name="l01279"></a><span class="lineno"> 1279</span>&#160;    <span class="comment">// test QSymmS16 quantization</span></div><div class="line"><a name="l01280"></a><span class="lineno"> 1280</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l01281"></a><span class="lineno"> 1281</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01282"></a><span class="lineno"> 1282</span>&#160;    TestInstanceNormalizationQuantization validatorQSymmS16(qSymmS16Options, tensorShape, tensorShape);</div><div class="line"><a name="l01283"></a><span class="lineno"> 1283</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l01284"></a><span class="lineno"> 1284</span>&#160;}</div><div class="line"><a name="l01285"></a><span class="lineno"> 1285</span>&#160;</div><div class="line"><a name="l01286"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a46d045b35ad6b8c2ffe0c04684f97779"> 1286</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeLogSoftmax)</div><div class="line"><a name="l01287"></a><span class="lineno"> 1287</span>&#160;{</div><div class="line"><a name="l01288"></a><span class="lineno"> 1288</span>&#160;    <span class="keyword">class </span>TestLogSoftmaxQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l01289"></a><span class="lineno"> 1289</span>&#160;    {</div><div class="line"><a name="l01290"></a><span class="lineno"> 1290</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l01291"></a><span class="lineno"> 1291</span>&#160;        TestLogSoftmaxQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01292"></a><span class="lineno"> 1292</span>&#160;            : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l01293"></a><span class="lineno"> 1293</span>&#160;</div><div class="line"><a name="l01294"></a><span class="lineno"> 1294</span>&#160;        TestLogSoftmaxQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l01295"></a><span class="lineno"> 1295</span>&#160;                                   <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l01296"></a><span class="lineno"> 1296</span>&#160;                                   <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01297"></a><span class="lineno"> 1297</span>&#160;            : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l01298"></a><span class="lineno"> 1298</span>&#160;</div><div class="line"><a name="l01299"></a><span class="lineno"> 1299</span>&#160;        <span class="keywordtype">void</span> VisitLogSoftmaxLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l01300"></a><span class="lineno"> 1300</span>&#160;                                  <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_softmax_descriptor.html">SoftmaxDescriptor</a>&amp; descriptor,</div><div class="line"><a name="l01301"></a><span class="lineno"> 1301</span>&#160;                                  <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01302"></a><span class="lineno"> 1302</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01303"></a><span class="lineno"> 1303</span>&#160;            boost::ignore_unused(descriptor, name);</div><div class="line"><a name="l01304"></a><span class="lineno"> 1304</span>&#160;            <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l01305"></a><span class="lineno"> 1305</span>&#160;</div><div class="line"><a name="l01306"></a><span class="lineno"> 1306</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qAsymmU8Params{ 30.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 128 };</div><div class="line"><a name="l01307"></a><span class="lineno"> 1307</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qAsymmS8Params { 30.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>,  0};</div><div class="line"><a name="l01308"></a><span class="lineno"> 1308</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qSymmS8Params { 15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0};</div><div class="line"><a name="l01309"></a><span class="lineno"> 1309</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qSymmS16Params{ 15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0 };</div><div class="line"><a name="l01310"></a><span class="lineno"> 1310</span>&#160;</div><div class="line"><a name="l01311"></a><span class="lineno"> 1311</span>&#160;            TestQuantizationParams(info, qAsymmU8Params, qAsymmS8Params, qSymmS8Params, qSymmS16Params);</div><div class="line"><a name="l01312"></a><span class="lineno"> 1312</span>&#160;        }</div><div class="line"><a name="l01313"></a><span class="lineno"> 1313</span>&#160;    };</div><div class="line"><a name="l01314"></a><span class="lineno"> 1314</span>&#160;</div><div class="line"><a name="l01315"></a><span class="lineno"> 1315</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> tensorShape{ 1U };</div><div class="line"><a name="l01316"></a><span class="lineno"> 1316</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> tensorInfo(tensorShape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l01317"></a><span class="lineno"> 1317</span>&#160;</div><div class="line"><a name="l01318"></a><span class="lineno"> 1318</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l01319"></a><span class="lineno"> 1319</span>&#160;</div><div class="line"><a name="l01320"></a><span class="lineno"> 1320</span>&#160;    <a class="code" href="structarmnn_1_1_softmax_descriptor.html">LogSoftmaxDescriptor</a> descriptor;</div><div class="line"><a name="l01321"></a><span class="lineno"> 1321</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_softmax_descriptor.html#a8275d51ef9a584feb95726ea0522f6e5">m_Beta</a> = 1.0f;</div><div class="line"><a name="l01322"></a><span class="lineno"> 1322</span>&#160;</div><div class="line"><a name="l01323"></a><span class="lineno"> 1323</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* inputLayer        = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l01324"></a><span class="lineno"> 1324</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* logSoftmaxLayer   = network-&gt;AddLogSoftmaxLayer(descriptor);</div><div class="line"><a name="l01325"></a><span class="lineno"> 1325</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* outputLayer       = network-&gt;AddOutputLayer(0);</div><div class="line"><a name="l01326"></a><span class="lineno"> 1326</span>&#160;</div><div class="line"><a name="l01327"></a><span class="lineno"> 1327</span>&#160;    inputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(logSoftmaxLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01328"></a><span class="lineno"> 1328</span>&#160;    logSoftmaxLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(outputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01329"></a><span class="lineno"> 1329</span>&#160;</div><div class="line"><a name="l01330"></a><span class="lineno"> 1330</span>&#160;    inputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(tensorInfo);</div><div class="line"><a name="l01331"></a><span class="lineno"> 1331</span>&#160;    logSoftmaxLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(tensorInfo);</div><div class="line"><a name="l01332"></a><span class="lineno"> 1332</span>&#160;</div><div class="line"><a name="l01333"></a><span class="lineno"> 1333</span>&#160;    <span class="comment">// test QAsymmU8 quantization</span></div><div class="line"><a name="l01334"></a><span class="lineno"> 1334</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l01335"></a><span class="lineno"> 1335</span>&#160;    TestLogSoftmaxQuantization validatorQAsymmU8(tensorShape, tensorShape);</div><div class="line"><a name="l01336"></a><span class="lineno"> 1336</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l01337"></a><span class="lineno"> 1337</span>&#160;</div><div class="line"><a name="l01338"></a><span class="lineno"> 1338</span>&#160;    <span class="comment">// test QAsymmS8 quantization</span></div><div class="line"><a name="l01339"></a><span class="lineno"> 1339</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l01340"></a><span class="lineno"> 1340</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01341"></a><span class="lineno"> 1341</span>&#160;    TestLogSoftmaxQuantization validatorQAsymmS8(qAsymmS8Options, tensorShape, tensorShape);</div><div class="line"><a name="l01342"></a><span class="lineno"> 1342</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l01343"></a><span class="lineno"> 1343</span>&#160;</div><div class="line"><a name="l01344"></a><span class="lineno"> 1344</span>&#160;    <span class="comment">// test QSymmS8 quantization</span></div><div class="line"><a name="l01345"></a><span class="lineno"> 1345</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l01346"></a><span class="lineno"> 1346</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01347"></a><span class="lineno"> 1347</span>&#160;    TestLogSoftmaxQuantization validatorQSymmS8(qSymmS8Options, tensorShape, tensorShape);</div><div class="line"><a name="l01348"></a><span class="lineno"> 1348</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l01349"></a><span class="lineno"> 1349</span>&#160;</div><div class="line"><a name="l01350"></a><span class="lineno"> 1350</span>&#160;    <span class="comment">// test QuantisedSymmS16 quantization</span></div><div class="line"><a name="l01351"></a><span class="lineno"> 1351</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l01352"></a><span class="lineno"> 1352</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l01353"></a><span class="lineno"> 1353</span>&#160;    TestLogSoftmaxQuantization validatorQSymmS16(qSymmS16options, tensorShape, tensorShape);</div><div class="line"><a name="l01354"></a><span class="lineno"> 1354</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l01355"></a><span class="lineno"> 1355</span>&#160;}</div><div class="line"><a name="l01356"></a><span class="lineno"> 1356</span>&#160;</div><div class="line"><a name="l01357"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a9c91b774c3089c55df77cc3a42da72de"> 1357</a></span>&#160;<a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> <a class="code" href="namespacearmnn.html#a9c91b774c3089c55df77cc3a42da72de">CreateNetworkWithSoftmaxLayer</a>(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_softmax_descriptor.html">SoftmaxDescriptor</a>&amp; descriptor, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; shape)</div><div class="line"><a name="l01358"></a><span class="lineno"> 1358</span>&#160;{</div><div class="line"><a name="l01359"></a><span class="lineno"> 1359</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l01360"></a><span class="lineno"> 1360</span>&#160;</div><div class="line"><a name="l01361"></a><span class="lineno"> 1361</span>&#160;    <span class="comment">// Add the layers</span></div><div class="line"><a name="l01362"></a><span class="lineno"> 1362</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input0 = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l01363"></a><span class="lineno"> 1363</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* softmax = network-&gt;AddSoftmaxLayer(descriptor);</div><div class="line"><a name="l01364"></a><span class="lineno"> 1364</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;AddOutputLayer(2);</div><div class="line"><a name="l01365"></a><span class="lineno"> 1365</span>&#160;</div><div class="line"><a name="l01366"></a><span class="lineno"> 1366</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l01367"></a><span class="lineno"> 1367</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(softmax-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01368"></a><span class="lineno"> 1368</span>&#160;    softmax-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01369"></a><span class="lineno"> 1369</span>&#160;</div><div class="line"><a name="l01370"></a><span class="lineno"> 1370</span>&#160;    <span class="comment">// Set TensorInfo</span></div><div class="line"><a name="l01371"></a><span class="lineno"> 1371</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l01372"></a><span class="lineno"> 1372</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l01373"></a><span class="lineno"> 1373</span>&#160;    softmax-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l01374"></a><span class="lineno"> 1374</span>&#160;</div><div class="line"><a name="l01375"></a><span class="lineno"> 1375</span>&#160;    <span class="keywordflow">return</span> network;</div><div class="line"><a name="l01376"></a><span class="lineno"> 1376</span>&#160;}</div><div class="line"><a name="l01377"></a><span class="lineno"> 1377</span>&#160;</div><div class="line"><a name="l01378"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a7e94e9ab356805c498f5fc2fba87e4e6"> 1378</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeSoftmax)</div><div class="line"><a name="l01379"></a><span class="lineno"> 1379</span>&#160;{</div><div class="line"><a name="l01380"></a><span class="lineno"> 1380</span>&#160;    <span class="keyword">class </span>TestSoftmaxQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l01381"></a><span class="lineno"> 1381</span>&#160;    {</div><div class="line"><a name="l01382"></a><span class="lineno"> 1382</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l01383"></a><span class="lineno"> 1383</span>&#160;        TestSoftmaxQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01384"></a><span class="lineno"> 1384</span>&#160;        : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l01385"></a><span class="lineno"> 1385</span>&#160;</div><div class="line"><a name="l01386"></a><span class="lineno"> 1386</span>&#160;        TestSoftmaxQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l01387"></a><span class="lineno"> 1387</span>&#160;                                <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l01388"></a><span class="lineno"> 1388</span>&#160;                                <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01389"></a><span class="lineno"> 1389</span>&#160;        : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l01390"></a><span class="lineno"> 1390</span>&#160;</div><div class="line"><a name="l01391"></a><span class="lineno"> 1391</span>&#160;        <span class="keywordtype">void</span> VisitSoftmaxLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l01392"></a><span class="lineno"> 1392</span>&#160;                               <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_softmax_descriptor.html">SoftmaxDescriptor</a>&amp; descriptor,</div><div class="line"><a name="l01393"></a><span class="lineno"> 1393</span>&#160;                               <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01394"></a><span class="lineno"> 1394</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01395"></a><span class="lineno"> 1395</span>&#160;            boost::ignore_unused(descriptor, name);</div><div class="line"><a name="l01396"></a><span class="lineno"> 1396</span>&#160;            <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l01397"></a><span class="lineno"> 1397</span>&#160;</div><div class="line"><a name="l01398"></a><span class="lineno"> 1398</span>&#160;            <span class="comment">// Based off default static range [0.0f, 1.0f]</span></div><div class="line"><a name="l01399"></a><span class="lineno"> 1399</span>&#160;            TestQuantizationParams(info, {1.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 0},</div><div class="line"><a name="l01400"></a><span class="lineno"> 1400</span>&#160;                                         {1.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>, -128},</div><div class="line"><a name="l01401"></a><span class="lineno"> 1401</span>&#160;                                         {1.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0},</div><div class="line"><a name="l01402"></a><span class="lineno"> 1402</span>&#160;                                         {1.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0});</div><div class="line"><a name="l01403"></a><span class="lineno"> 1403</span>&#160;        }</div><div class="line"><a name="l01404"></a><span class="lineno"> 1404</span>&#160;    };</div><div class="line"><a name="l01405"></a><span class="lineno"> 1405</span>&#160;</div><div class="line"><a name="l01406"></a><span class="lineno"> 1406</span>&#160;    <a class="code" href="structarmnn_1_1_softmax_descriptor.html">SoftmaxDescriptor</a> descriptor;</div><div class="line"><a name="l01407"></a><span class="lineno"> 1407</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_softmax_descriptor.html#a8275d51ef9a584feb95726ea0522f6e5">m_Beta</a> = 1.0f;</div><div class="line"><a name="l01408"></a><span class="lineno"> 1408</span>&#160;</div><div class="line"><a name="l01409"></a><span class="lineno"> 1409</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l01410"></a><span class="lineno"> 1410</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="namespacearmnn.html#a9c91b774c3089c55df77cc3a42da72de">CreateNetworkWithSoftmaxLayer</a>(descriptor, shape);</div><div class="line"><a name="l01411"></a><span class="lineno"> 1411</span>&#160;</div><div class="line"><a name="l01412"></a><span class="lineno"> 1412</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l01413"></a><span class="lineno"> 1413</span>&#160;    TestSoftmaxQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l01414"></a><span class="lineno"> 1414</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l01415"></a><span class="lineno"> 1415</span>&#160;</div><div class="line"><a name="l01416"></a><span class="lineno"> 1416</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l01417"></a><span class="lineno"> 1417</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01418"></a><span class="lineno"> 1418</span>&#160;    TestSoftmaxQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l01419"></a><span class="lineno"> 1419</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l01420"></a><span class="lineno"> 1420</span>&#160;</div><div class="line"><a name="l01421"></a><span class="lineno"> 1421</span>&#160;    <span class="comment">// test QSymmS8 quantization</span></div><div class="line"><a name="l01422"></a><span class="lineno"> 1422</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l01423"></a><span class="lineno"> 1423</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01424"></a><span class="lineno"> 1424</span>&#160;    TestSoftmaxQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l01425"></a><span class="lineno"> 1425</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l01426"></a><span class="lineno"> 1426</span>&#160;</div><div class="line"><a name="l01427"></a><span class="lineno"> 1427</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l01428"></a><span class="lineno"> 1428</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l01429"></a><span class="lineno"> 1429</span>&#160;    TestSoftmaxQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l01430"></a><span class="lineno"> 1430</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l01431"></a><span class="lineno"> 1431</span>&#160;}</div><div class="line"><a name="l01432"></a><span class="lineno"> 1432</span>&#160;</div><div class="line"><a name="l01433"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a4734542212b5811d0511ea6b16f35168"> 1433</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeStandIn)</div><div class="line"><a name="l01434"></a><span class="lineno"> 1434</span>&#160;{</div><div class="line"><a name="l01435"></a><span class="lineno"> 1435</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> tensorShape{ 1U };</div><div class="line"><a name="l01436"></a><span class="lineno"> 1436</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> tensorInfo(tensorShape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l01437"></a><span class="lineno"> 1437</span>&#160;</div><div class="line"><a name="l01438"></a><span class="lineno"> 1438</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l01439"></a><span class="lineno"> 1439</span>&#160;</div><div class="line"><a name="l01440"></a><span class="lineno"> 1440</span>&#160;    <a class="code" href="structarmnn_1_1_stand_in_descriptor.html">StandInDescriptor</a> descriptor;</div><div class="line"><a name="l01441"></a><span class="lineno"> 1441</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_stand_in_descriptor.html#aed6086070440ceb94129bef06f70173f">m_NumInputs</a> = 1;</div><div class="line"><a name="l01442"></a><span class="lineno"> 1442</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_stand_in_descriptor.html#abb8a2d2bb8cc594c26aaa70c820ac5cc">m_NumOutputs</a> = 1;</div><div class="line"><a name="l01443"></a><span class="lineno"> 1443</span>&#160;</div><div class="line"><a name="l01444"></a><span class="lineno"> 1444</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* inputLayer     = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l01445"></a><span class="lineno"> 1445</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* standInLayer   = network-&gt;AddStandInLayer(descriptor);</div><div class="line"><a name="l01446"></a><span class="lineno"> 1446</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* outputLayer    = network-&gt;AddOutputLayer(0);</div><div class="line"><a name="l01447"></a><span class="lineno"> 1447</span>&#160;</div><div class="line"><a name="l01448"></a><span class="lineno"> 1448</span>&#160;    inputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(standInLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01449"></a><span class="lineno"> 1449</span>&#160;    standInLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(outputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01450"></a><span class="lineno"> 1450</span>&#160;</div><div class="line"><a name="l01451"></a><span class="lineno"> 1451</span>&#160;    inputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(tensorInfo);</div><div class="line"><a name="l01452"></a><span class="lineno"> 1452</span>&#160;    standInLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(tensorInfo);</div><div class="line"><a name="l01453"></a><span class="lineno"> 1453</span>&#160;</div><div class="line"><a name="l01454"></a><span class="lineno"> 1454</span>&#160;    <span class="comment">// test QAsymmU8 quantization</span></div><div class="line"><a name="l01455"></a><span class="lineno"> 1455</span>&#160;    BOOST_CHECK_THROW(<a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork(),</div><div class="line"><a name="l01456"></a><span class="lineno"> 1456</span>&#160;                      <a class="code" href="classarmnn_1_1_unimplemented_exception.html">armnn::UnimplementedException</a>);</div><div class="line"><a name="l01457"></a><span class="lineno"> 1457</span>&#160;</div><div class="line"><a name="l01458"></a><span class="lineno"> 1458</span>&#160;    <span class="comment">// test QAsymmS8 quantization</span></div><div class="line"><a name="l01459"></a><span class="lineno"> 1459</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l01460"></a><span class="lineno"> 1460</span>&#160;    BOOST_CHECK_THROW(<a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork(),</div><div class="line"><a name="l01461"></a><span class="lineno"> 1461</span>&#160;                      <a class="code" href="classarmnn_1_1_unimplemented_exception.html">armnn::UnimplementedException</a>);</div><div class="line"><a name="l01462"></a><span class="lineno"> 1462</span>&#160;</div><div class="line"><a name="l01463"></a><span class="lineno"> 1463</span>&#160;    <span class="comment">// test QuantisedSymmS16 quantization</span></div><div class="line"><a name="l01464"></a><span class="lineno"> 1464</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l01465"></a><span class="lineno"> 1465</span>&#160;    BOOST_CHECK_THROW(<a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork(),</div><div class="line"><a name="l01466"></a><span class="lineno"> 1466</span>&#160;                      <a class="code" href="classarmnn_1_1_unimplemented_exception.html">armnn::UnimplementedException</a>);</div><div class="line"><a name="l01467"></a><span class="lineno"> 1467</span>&#160;</div><div class="line"><a name="l01468"></a><span class="lineno"> 1468</span>&#160;    <span class="comment">// test QuantisedSymmS16 quantization</span></div><div class="line"><a name="l01469"></a><span class="lineno"> 1469</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l01470"></a><span class="lineno"> 1470</span>&#160;    BOOST_CHECK_THROW(<a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork(),</div><div class="line"><a name="l01471"></a><span class="lineno"> 1471</span>&#160;                      <a class="code" href="classarmnn_1_1_unimplemented_exception.html">armnn::UnimplementedException</a>);</div><div class="line"><a name="l01472"></a><span class="lineno"> 1472</span>&#160;}</div><div class="line"><a name="l01473"></a><span class="lineno"> 1473</span>&#160;</div><div class="line"><a name="l01474"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a120c131df35d78b3a56cb0f07decaf35"> 1474</a></span>&#160;<a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* <a class="code" href="namespacearmnn.html#a120c131df35d78b3a56cb0f07decaf35">CreateStartOfLeakyReluNetwork</a>(<a class="code" href="classarmnn_1_1_i_network.html">INetwork</a>* network, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>)</div><div class="line"><a name="l01475"></a><span class="lineno"> 1475</span>&#160;{</div><div class="line"><a name="l01476"></a><span class="lineno"> 1476</span>&#160;    <a class="code" href="structarmnn_1_1_activation_descriptor.html">ActivationDescriptor</a> activationDescriptor;</div><div class="line"><a name="l01477"></a><span class="lineno"> 1477</span>&#160;    activationDescriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#af10fa7883e3579950f477bee92a64844">m_Function</a> = <a class="code" href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaacb7667f5ec2f6e8a5943b781ba6c2735">ActivationFunction::LeakyReLu</a>;</div><div class="line"><a name="l01478"></a><span class="lineno"> 1478</span>&#160;    activationDescriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a017b2990003a014234f13e999dc7c689">m_A</a>        = 3.5f;</div><div class="line"><a name="l01479"></a><span class="lineno"> 1479</span>&#160;    activationDescriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a28c4c9cb15f6be3499abbc46b356060b">m_B</a>        = -10.0f;</div><div class="line"><a name="l01480"></a><span class="lineno"> 1480</span>&#160;</div><div class="line"><a name="l01481"></a><span class="lineno"> 1481</span>&#160;    <span class="comment">// Add the layers</span></div><div class="line"><a name="l01482"></a><span class="lineno"> 1482</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input0 = network-&gt;<a class="code" href="classarmnn_1_1_i_network.html#a87d5ec72def73ca14bd2987a024bd569">AddInputLayer</a>(0);</div><div class="line"><a name="l01483"></a><span class="lineno"> 1483</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* activation = network-&gt;<a class="code" href="classarmnn_1_1_i_network.html#aea068f6094e1c3bfcdf8167b68112632">AddActivationLayer</a>(activationDescriptor);</div><div class="line"><a name="l01484"></a><span class="lineno"> 1484</span>&#160;</div><div class="line"><a name="l01485"></a><span class="lineno"> 1485</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l01486"></a><span class="lineno"> 1486</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(activation-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01487"></a><span class="lineno"> 1487</span>&#160;</div><div class="line"><a name="l01488"></a><span class="lineno"> 1488</span>&#160;    <span class="comment">// Set TensorInfo</span></div><div class="line"><a name="l01489"></a><span class="lineno"> 1489</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l01490"></a><span class="lineno"> 1490</span>&#160;    activation-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l01491"></a><span class="lineno"> 1491</span>&#160;</div><div class="line"><a name="l01492"></a><span class="lineno"> 1492</span>&#160;    <span class="keywordflow">return</span> activation;</div><div class="line"><a name="l01493"></a><span class="lineno"> 1493</span>&#160;}</div><div class="line"><a name="l01494"></a><span class="lineno"> 1494</span>&#160;</div><div class="line"><a name="l01495"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a6fff4b4b1b5d4d37c9cf53d0e31c05dd"> 1495</a></span>&#160;<span class="keywordtype">void</span>  <a class="code" href="namespacearmnn.html#a6fff4b4b1b5d4d37c9cf53d0e31c05dd">CompleteLeakyReluNetwork</a>(<a class="code" href="classarmnn_1_1_i_network.html">INetwork</a>* network,</div><div class="line"><a name="l01496"></a><span class="lineno"> 1496</span>&#160;                               <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* activation,</div><div class="line"><a name="l01497"></a><span class="lineno"> 1497</span>&#160;                               <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layerUnderTest,</div><div class="line"><a name="l01498"></a><span class="lineno"> 1498</span>&#160;                               <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>)</div><div class="line"><a name="l01499"></a><span class="lineno"> 1499</span>&#160;{</div><div class="line"><a name="l01500"></a><span class="lineno"> 1500</span>&#160;    <span class="comment">// Add the output Layer</span></div><div class="line"><a name="l01501"></a><span class="lineno"> 1501</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;<a class="code" href="classarmnn_1_1_i_network.html#ad8582fba2ebeb65da43a56bc22d4f88b">AddOutputLayer</a>(3);</div><div class="line"><a name="l01502"></a><span class="lineno"> 1502</span>&#160;</div><div class="line"><a name="l01503"></a><span class="lineno"> 1503</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l01504"></a><span class="lineno"> 1504</span>&#160;    activation-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(layerUnderTest-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01505"></a><span class="lineno"> 1505</span>&#160;    layerUnderTest-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01506"></a><span class="lineno"> 1506</span>&#160;</div><div class="line"><a name="l01507"></a><span class="lineno"> 1507</span>&#160;    <span class="comment">//Set TensorInfo</span></div><div class="line"><a name="l01508"></a><span class="lineno"> 1508</span>&#160;    layerUnderTest-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l01509"></a><span class="lineno"> 1509</span>&#160;}</div><div class="line"><a name="l01510"></a><span class="lineno"> 1510</span>&#160;</div><div class="line"><a name="l01511"></a><span class="lineno"><a class="line" href="namespacearmnn.html#add22da50dd35a100548dde4c57ae89d1"> 1511</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizePermute)</div><div class="line"><a name="l01512"></a><span class="lineno"> 1512</span>&#160;{</div><div class="line"><a name="l01513"></a><span class="lineno"> 1513</span>&#160;    <span class="keyword">class </span>TestPermuteQuantization : <span class="keyword">public</span> TestLeakyReLuActivationQuantization</div><div class="line"><a name="l01514"></a><span class="lineno"> 1514</span>&#160;    {</div><div class="line"><a name="l01515"></a><span class="lineno"> 1515</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l01516"></a><span class="lineno"> 1516</span>&#160;        TestPermuteQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01517"></a><span class="lineno"> 1517</span>&#160;        : TestLeakyReLuActivationQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l01518"></a><span class="lineno"> 1518</span>&#160;</div><div class="line"><a name="l01519"></a><span class="lineno"> 1519</span>&#160;        TestPermuteQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l01520"></a><span class="lineno"> 1520</span>&#160;                                <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l01521"></a><span class="lineno"> 1521</span>&#160;                                <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01522"></a><span class="lineno"> 1522</span>&#160;        : TestLeakyReLuActivationQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l01523"></a><span class="lineno"> 1523</span>&#160;</div><div class="line"><a name="l01524"></a><span class="lineno"> 1524</span>&#160;        <span class="keywordtype">void</span> VisitPermuteLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l01525"></a><span class="lineno"> 1525</span>&#160;                               <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_permute_descriptor.html">PermuteDescriptor</a>&amp; desc,</div><div class="line"><a name="l01526"></a><span class="lineno"> 1526</span>&#160;                               <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01527"></a><span class="lineno"> 1527</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01528"></a><span class="lineno"> 1528</span>&#160;            boost::ignore_unused(desc, name);</div><div class="line"><a name="l01529"></a><span class="lineno"> 1529</span>&#160;            CheckForwardedQuantizationSettings(layer);</div><div class="line"><a name="l01530"></a><span class="lineno"> 1530</span>&#160;        }</div><div class="line"><a name="l01531"></a><span class="lineno"> 1531</span>&#160;    };</div><div class="line"><a name="l01532"></a><span class="lineno"> 1532</span>&#160;</div><div class="line"><a name="l01533"></a><span class="lineno"> 1533</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l01534"></a><span class="lineno"> 1534</span>&#160;</div><div class="line"><a name="l01535"></a><span class="lineno"> 1535</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l01536"></a><span class="lineno"> 1536</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l01537"></a><span class="lineno"> 1537</span>&#160;</div><div class="line"><a name="l01538"></a><span class="lineno"> 1538</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* activation = <a class="code" href="namespacearmnn.html#a120c131df35d78b3a56cb0f07decaf35">CreateStartOfLeakyReluNetwork</a>(network.get(), <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l01539"></a><span class="lineno"> 1539</span>&#160;</div><div class="line"><a name="l01540"></a><span class="lineno"> 1540</span>&#160;    <span class="comment">// Add the layer under test</span></div><div class="line"><a name="l01541"></a><span class="lineno"> 1541</span>&#160;    <a class="code" href="structarmnn_1_1_permute_descriptor.html">PermuteDescriptor</a> desc;</div><div class="line"><a name="l01542"></a><span class="lineno"> 1542</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* permute = network-&gt;AddPermuteLayer(desc);</div><div class="line"><a name="l01543"></a><span class="lineno"> 1543</span>&#160;</div><div class="line"><a name="l01544"></a><span class="lineno"> 1544</span>&#160;    <a class="code" href="namespacearmnn.html#a6fff4b4b1b5d4d37c9cf53d0e31c05dd">CompleteLeakyReluNetwork</a>(network.get(), activation, permute, <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l01545"></a><span class="lineno"> 1545</span>&#160;</div><div class="line"><a name="l01546"></a><span class="lineno"> 1546</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l01547"></a><span class="lineno"> 1547</span>&#160;    TestPermuteQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l01548"></a><span class="lineno"> 1548</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l01549"></a><span class="lineno"> 1549</span>&#160;</div><div class="line"><a name="l01550"></a><span class="lineno"> 1550</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l01551"></a><span class="lineno"> 1551</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01552"></a><span class="lineno"> 1552</span>&#160;    TestPermuteQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l01553"></a><span class="lineno"> 1553</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l01554"></a><span class="lineno"> 1554</span>&#160;</div><div class="line"><a name="l01555"></a><span class="lineno"> 1555</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l01556"></a><span class="lineno"> 1556</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01557"></a><span class="lineno"> 1557</span>&#160;    TestPermuteQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l01558"></a><span class="lineno"> 1558</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l01559"></a><span class="lineno"> 1559</span>&#160;</div><div class="line"><a name="l01560"></a><span class="lineno"> 1560</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l01561"></a><span class="lineno"> 1561</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l01562"></a><span class="lineno"> 1562</span>&#160;    TestPermuteQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l01563"></a><span class="lineno"> 1563</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l01564"></a><span class="lineno"> 1564</span>&#160;}</div><div class="line"><a name="l01565"></a><span class="lineno"> 1565</span>&#160;</div><div class="line"><a name="l01566"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a9a6bc66017eb7c132fd6e13ff0dcb540"> 1566</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeSpaceToBatch)</div><div class="line"><a name="l01567"></a><span class="lineno"> 1567</span>&#160;{</div><div class="line"><a name="l01568"></a><span class="lineno"> 1568</span>&#160;    <span class="keyword">class </span>TestSpaceToBatchQuantization : <span class="keyword">public</span> TestLeakyReLuActivationQuantization</div><div class="line"><a name="l01569"></a><span class="lineno"> 1569</span>&#160;    {</div><div class="line"><a name="l01570"></a><span class="lineno"> 1570</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l01571"></a><span class="lineno"> 1571</span>&#160;        TestSpaceToBatchQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01572"></a><span class="lineno"> 1572</span>&#160;        : TestLeakyReLuActivationQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l01573"></a><span class="lineno"> 1573</span>&#160;</div><div class="line"><a name="l01574"></a><span class="lineno"> 1574</span>&#160;        TestSpaceToBatchQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l01575"></a><span class="lineno"> 1575</span>&#160;                                     <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l01576"></a><span class="lineno"> 1576</span>&#160;                                     <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01577"></a><span class="lineno"> 1577</span>&#160;        : TestLeakyReLuActivationQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l01578"></a><span class="lineno"> 1578</span>&#160;</div><div class="line"><a name="l01579"></a><span class="lineno"> 1579</span>&#160;        <span class="keywordtype">void</span> VisitSpaceToBatchNdLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l01580"></a><span class="lineno"> 1580</span>&#160;                                      <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_space_to_batch_nd_descriptor.html">SpaceToBatchNdDescriptor</a>&amp; spaceToBatchNdDescriptor,</div><div class="line"><a name="l01581"></a><span class="lineno"> 1581</span>&#160;                                      <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01582"></a><span class="lineno"> 1582</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01583"></a><span class="lineno"> 1583</span>&#160;            boost::ignore_unused(spaceToBatchNdDescriptor, name);</div><div class="line"><a name="l01584"></a><span class="lineno"> 1584</span>&#160;            CheckForwardedQuantizationSettings(layer);</div><div class="line"><a name="l01585"></a><span class="lineno"> 1585</span>&#160;        }</div><div class="line"><a name="l01586"></a><span class="lineno"> 1586</span>&#160;    };</div><div class="line"><a name="l01587"></a><span class="lineno"> 1587</span>&#160;</div><div class="line"><a name="l01588"></a><span class="lineno"> 1588</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l01589"></a><span class="lineno"> 1589</span>&#160;</div><div class="line"><a name="l01590"></a><span class="lineno"> 1590</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l01591"></a><span class="lineno"> 1591</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l01592"></a><span class="lineno"> 1592</span>&#160;</div><div class="line"><a name="l01593"></a><span class="lineno"> 1593</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* activation = <a class="code" href="namespacearmnn.html#a120c131df35d78b3a56cb0f07decaf35">CreateStartOfLeakyReluNetwork</a>(network.get(), <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l01594"></a><span class="lineno"> 1594</span>&#160;</div><div class="line"><a name="l01595"></a><span class="lineno"> 1595</span>&#160;    <span class="comment">// Add the layer under test</span></div><div class="line"><a name="l01596"></a><span class="lineno"> 1596</span>&#160;    <a class="code" href="structarmnn_1_1_space_to_batch_nd_descriptor.html">SpaceToBatchNdDescriptor</a> descriptor;</div><div class="line"><a name="l01597"></a><span class="lineno"> 1597</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* spaceToBatch = network-&gt;AddSpaceToBatchNdLayer(descriptor);</div><div class="line"><a name="l01598"></a><span class="lineno"> 1598</span>&#160;</div><div class="line"><a name="l01599"></a><span class="lineno"> 1599</span>&#160;    <a class="code" href="namespacearmnn.html#a6fff4b4b1b5d4d37c9cf53d0e31c05dd">CompleteLeakyReluNetwork</a>(network.get(), activation, spaceToBatch, <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l01600"></a><span class="lineno"> 1600</span>&#160;</div><div class="line"><a name="l01601"></a><span class="lineno"> 1601</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l01602"></a><span class="lineno"> 1602</span>&#160;    TestSpaceToBatchQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l01603"></a><span class="lineno"> 1603</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l01604"></a><span class="lineno"> 1604</span>&#160;</div><div class="line"><a name="l01605"></a><span class="lineno"> 1605</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l01606"></a><span class="lineno"> 1606</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01607"></a><span class="lineno"> 1607</span>&#160;    TestSpaceToBatchQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l01608"></a><span class="lineno"> 1608</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l01609"></a><span class="lineno"> 1609</span>&#160;</div><div class="line"><a name="l01610"></a><span class="lineno"> 1610</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l01611"></a><span class="lineno"> 1611</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01612"></a><span class="lineno"> 1612</span>&#160;    TestSpaceToBatchQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l01613"></a><span class="lineno"> 1613</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l01614"></a><span class="lineno"> 1614</span>&#160;</div><div class="line"><a name="l01615"></a><span class="lineno"> 1615</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l01616"></a><span class="lineno"> 1616</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l01617"></a><span class="lineno"> 1617</span>&#160;    TestSpaceToBatchQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l01618"></a><span class="lineno"> 1618</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l01619"></a><span class="lineno"> 1619</span>&#160;}</div><div class="line"><a name="l01620"></a><span class="lineno"> 1620</span>&#160;</div><div class="line"><a name="l01621"></a><span class="lineno"><a class="line" href="namespacearmnn.html#aa78ce2bbe65cae8f3d60839467dbfc83"> 1621</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeSpaceToDepth)</div><div class="line"><a name="l01622"></a><span class="lineno"> 1622</span>&#160;{</div><div class="line"><a name="l01623"></a><span class="lineno"> 1623</span>&#160;    <span class="keyword">class </span>TestSpaceToDepthQuantization : <span class="keyword">public</span> TestLeakyReLuActivationQuantization</div><div class="line"><a name="l01624"></a><span class="lineno"> 1624</span>&#160;    {</div><div class="line"><a name="l01625"></a><span class="lineno"> 1625</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l01626"></a><span class="lineno"> 1626</span>&#160;        TestSpaceToDepthQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01627"></a><span class="lineno"> 1627</span>&#160;            : TestLeakyReLuActivationQuantization(inputShape, outputShape)</div><div class="line"><a name="l01628"></a><span class="lineno"> 1628</span>&#160;        {}</div><div class="line"><a name="l01629"></a><span class="lineno"> 1629</span>&#160;</div><div class="line"><a name="l01630"></a><span class="lineno"> 1630</span>&#160;        TestSpaceToDepthQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l01631"></a><span class="lineno"> 1631</span>&#160;                                     <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l01632"></a><span class="lineno"> 1632</span>&#160;                                     <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01633"></a><span class="lineno"> 1633</span>&#160;            : TestLeakyReLuActivationQuantization(options, inputShape, outputShape)</div><div class="line"><a name="l01634"></a><span class="lineno"> 1634</span>&#160;        {}</div><div class="line"><a name="l01635"></a><span class="lineno"> 1635</span>&#160;</div><div class="line"><a name="l01636"></a><span class="lineno"> 1636</span>&#160;        <span class="keywordtype">void</span> VisitSpaceToDepthLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l01637"></a><span class="lineno"> 1637</span>&#160;                                    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_space_to_depth_descriptor.html">SpaceToDepthDescriptor</a>&amp;,</div><div class="line"><a name="l01638"></a><span class="lineno"> 1638</span>&#160;                                    <span class="keyword">const</span> <span class="keywordtype">char</span>* = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01639"></a><span class="lineno"> 1639</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01640"></a><span class="lineno"> 1640</span>&#160;            <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l01641"></a><span class="lineno"> 1641</span>&#160;            TestQuantizationParams(info,</div><div class="line"><a name="l01642"></a><span class="lineno"> 1642</span>&#160;                                  { 30.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 128 },</div><div class="line"><a name="l01643"></a><span class="lineno"> 1643</span>&#160;                                  { 30.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>, 0   },</div><div class="line"><a name="l01644"></a><span class="lineno"> 1644</span>&#160;                                  { 15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0   },</div><div class="line"><a name="l01645"></a><span class="lineno"> 1645</span>&#160;                                  { 15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0   });</div><div class="line"><a name="l01646"></a><span class="lineno"> 1646</span>&#160;        }</div><div class="line"><a name="l01647"></a><span class="lineno"> 1647</span>&#160;    };</div><div class="line"><a name="l01648"></a><span class="lineno"> 1648</span>&#160;</div><div class="line"><a name="l01649"></a><span class="lineno"> 1649</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l01650"></a><span class="lineno"> 1650</span>&#160;</div><div class="line"><a name="l01651"></a><span class="lineno"> 1651</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{ 1u };</div><div class="line"><a name="l01652"></a><span class="lineno"> 1652</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l01653"></a><span class="lineno"> 1653</span>&#160;</div><div class="line"><a name="l01654"></a><span class="lineno"> 1654</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* activation   = <a class="code" href="namespacearmnn.html#a120c131df35d78b3a56cb0f07decaf35">CreateStartOfLeakyReluNetwork</a>(network.get(), <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l01655"></a><span class="lineno"> 1655</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* spaceToDepth = network-&gt;AddSpaceToDepthLayer(<a class="code" href="structarmnn_1_1_space_to_depth_descriptor.html">SpaceToDepthDescriptor</a>());</div><div class="line"><a name="l01656"></a><span class="lineno"> 1656</span>&#160;</div><div class="line"><a name="l01657"></a><span class="lineno"> 1657</span>&#160;    <a class="code" href="namespacearmnn.html#a6fff4b4b1b5d4d37c9cf53d0e31c05dd">CompleteLeakyReluNetwork</a>(network.get(), activation, spaceToDepth, <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l01658"></a><span class="lineno"> 1658</span>&#160;</div><div class="line"><a name="l01659"></a><span class="lineno"> 1659</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l01660"></a><span class="lineno"> 1660</span>&#160;    TestSpaceToDepthQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l01661"></a><span class="lineno"> 1661</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l01662"></a><span class="lineno"> 1662</span>&#160;</div><div class="line"><a name="l01663"></a><span class="lineno"> 1663</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l01664"></a><span class="lineno"> 1664</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01665"></a><span class="lineno"> 1665</span>&#160;    TestSpaceToDepthQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l01666"></a><span class="lineno"> 1666</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l01667"></a><span class="lineno"> 1667</span>&#160;</div><div class="line"><a name="l01668"></a><span class="lineno"> 1668</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l01669"></a><span class="lineno"> 1669</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01670"></a><span class="lineno"> 1670</span>&#160;    TestSpaceToDepthQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l01671"></a><span class="lineno"> 1671</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l01672"></a><span class="lineno"> 1672</span>&#160;</div><div class="line"><a name="l01673"></a><span class="lineno"> 1673</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l01674"></a><span class="lineno"> 1674</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l01675"></a><span class="lineno"> 1675</span>&#160;    TestSpaceToDepthQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l01676"></a><span class="lineno"> 1676</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l01677"></a><span class="lineno"> 1677</span>&#160;}</div><div class="line"><a name="l01678"></a><span class="lineno"> 1678</span>&#160;</div><div class="line"><a name="l01679"></a><span class="lineno"><a class="line" href="namespacearmnn.html#aaa86b6903e41d2d2828e00b32f872375"> 1679</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizePooling2d)</div><div class="line"><a name="l01680"></a><span class="lineno"> 1680</span>&#160;{</div><div class="line"><a name="l01681"></a><span class="lineno"> 1681</span>&#160;    <span class="keyword">class </span>TestPooling2dQuantization : <span class="keyword">public</span> TestLeakyReLuActivationQuantization</div><div class="line"><a name="l01682"></a><span class="lineno"> 1682</span>&#160;    {</div><div class="line"><a name="l01683"></a><span class="lineno"> 1683</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l01684"></a><span class="lineno"> 1684</span>&#160;        TestPooling2dQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01685"></a><span class="lineno"> 1685</span>&#160;        : TestLeakyReLuActivationQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l01686"></a><span class="lineno"> 1686</span>&#160;</div><div class="line"><a name="l01687"></a><span class="lineno"> 1687</span>&#160;        TestPooling2dQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l01688"></a><span class="lineno"> 1688</span>&#160;                                  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l01689"></a><span class="lineno"> 1689</span>&#160;                                  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01690"></a><span class="lineno"> 1690</span>&#160;        : TestLeakyReLuActivationQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l01691"></a><span class="lineno"> 1691</span>&#160;</div><div class="line"><a name="l01692"></a><span class="lineno"> 1692</span>&#160;        <span class="keywordtype">void</span> VisitPooling2dLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l01693"></a><span class="lineno"> 1693</span>&#160;                                 <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_pooling2d_descriptor.html">Pooling2dDescriptor</a>&amp; desc,</div><div class="line"><a name="l01694"></a><span class="lineno"> 1694</span>&#160;                                 <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01695"></a><span class="lineno"> 1695</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01696"></a><span class="lineno"> 1696</span>&#160;            boost::ignore_unused(desc, name);</div><div class="line"><a name="l01697"></a><span class="lineno"> 1697</span>&#160;            CheckForwardedQuantizationSettings(layer);</div><div class="line"><a name="l01698"></a><span class="lineno"> 1698</span>&#160;        }</div><div class="line"><a name="l01699"></a><span class="lineno"> 1699</span>&#160;    };</div><div class="line"><a name="l01700"></a><span class="lineno"> 1700</span>&#160;</div><div class="line"><a name="l01701"></a><span class="lineno"> 1701</span>&#160;    <span class="keyword">auto</span> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l01702"></a><span class="lineno"> 1702</span>&#160;</div><div class="line"><a name="l01703"></a><span class="lineno"> 1703</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l01704"></a><span class="lineno"> 1704</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l01705"></a><span class="lineno"> 1705</span>&#160;</div><div class="line"><a name="l01706"></a><span class="lineno"> 1706</span>&#160;    <a class="code" href="structarmnn_1_1_pooling2d_descriptor.html">Pooling2dDescriptor</a> desc;</div><div class="line"><a name="l01707"></a><span class="lineno"> 1707</span>&#160;    <a class="code" href="structarmnn_1_1_activation_descriptor.html">ActivationDescriptor</a> activationDescriptor;</div><div class="line"><a name="l01708"></a><span class="lineno"> 1708</span>&#160;    activationDescriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#af10fa7883e3579950f477bee92a64844">m_Function</a> = <a class="code" href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaacb7667f5ec2f6e8a5943b781ba6c2735">ActivationFunction::LeakyReLu</a>;</div><div class="line"><a name="l01709"></a><span class="lineno"> 1709</span>&#160;    activationDescriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a017b2990003a014234f13e999dc7c689">m_A</a>        = 3.5f;</div><div class="line"><a name="l01710"></a><span class="lineno"> 1710</span>&#160;    activationDescriptor.<a class="code" href="structarmnn_1_1_activation_descriptor.html#a28c4c9cb15f6be3499abbc46b356060b">m_B</a>        = -10.0f;</div><div class="line"><a name="l01711"></a><span class="lineno"> 1711</span>&#160;</div><div class="line"><a name="l01712"></a><span class="lineno"> 1712</span>&#160;    <span class="comment">// Add the layers</span></div><div class="line"><a name="l01713"></a><span class="lineno"> 1713</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input0 = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l01714"></a><span class="lineno"> 1714</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* activation = network-&gt;AddActivationLayer(activationDescriptor);</div><div class="line"><a name="l01715"></a><span class="lineno"> 1715</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* pooling2d = network-&gt;AddPooling2dLayer(desc);</div><div class="line"><a name="l01716"></a><span class="lineno"> 1716</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;AddOutputLayer(3);</div><div class="line"><a name="l01717"></a><span class="lineno"> 1717</span>&#160;</div><div class="line"><a name="l01718"></a><span class="lineno"> 1718</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l01719"></a><span class="lineno"> 1719</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(activation-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01720"></a><span class="lineno"> 1720</span>&#160;    activation-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(pooling2d-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01721"></a><span class="lineno"> 1721</span>&#160;    pooling2d-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01722"></a><span class="lineno"> 1722</span>&#160;</div><div class="line"><a name="l01723"></a><span class="lineno"> 1723</span>&#160;    <span class="comment">// Set TensorInfo</span></div><div class="line"><a name="l01724"></a><span class="lineno"> 1724</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l01725"></a><span class="lineno"> 1725</span>&#160;    activation-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l01726"></a><span class="lineno"> 1726</span>&#160;    pooling2d-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l01727"></a><span class="lineno"> 1727</span>&#160;</div><div class="line"><a name="l01728"></a><span class="lineno"> 1728</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l01729"></a><span class="lineno"> 1729</span>&#160;    TestPooling2dQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l01730"></a><span class="lineno"> 1730</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l01731"></a><span class="lineno"> 1731</span>&#160;</div><div class="line"><a name="l01732"></a><span class="lineno"> 1732</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l01733"></a><span class="lineno"> 1733</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01734"></a><span class="lineno"> 1734</span>&#160;    TestPooling2dQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l01735"></a><span class="lineno"> 1735</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l01736"></a><span class="lineno"> 1736</span>&#160;</div><div class="line"><a name="l01737"></a><span class="lineno"> 1737</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l01738"></a><span class="lineno"> 1738</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01739"></a><span class="lineno"> 1739</span>&#160;    TestPooling2dQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l01740"></a><span class="lineno"> 1740</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l01741"></a><span class="lineno"> 1741</span>&#160;</div><div class="line"><a name="l01742"></a><span class="lineno"> 1742</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l01743"></a><span class="lineno"> 1743</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l01744"></a><span class="lineno"> 1744</span>&#160;    TestPooling2dQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l01745"></a><span class="lineno"> 1745</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l01746"></a><span class="lineno"> 1746</span>&#160;}</div><div class="line"><a name="l01747"></a><span class="lineno"> 1747</span>&#160;</div><div class="line"><a name="l01748"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a369051e180246c66b20c93de5fecee8c"> 1748</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(<a class="code" href="namespacearmnn.html#a0e2bce68a1f7eff47ead4d9a2804eb91">QuantizeConstant</a>)</div><div class="line"><a name="l01749"></a><span class="lineno"> 1749</span>&#160;{</div><div class="line"><a name="l01750"></a><span class="lineno"> 1750</span>&#160;    <span class="keyword">class </span>TestConstantQuantization : <span class="keyword">public</span> TestAdditionQuantization</div><div class="line"><a name="l01751"></a><span class="lineno"> 1751</span>&#160;    {</div><div class="line"><a name="l01752"></a><span class="lineno"> 1752</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l01753"></a><span class="lineno"> 1753</span>&#160;        TestConstantQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01754"></a><span class="lineno"> 1754</span>&#160;        : TestAdditionQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l01755"></a><span class="lineno"> 1755</span>&#160;</div><div class="line"><a name="l01756"></a><span class="lineno"> 1756</span>&#160;        TestConstantQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l01757"></a><span class="lineno"> 1757</span>&#160;                                 <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l01758"></a><span class="lineno"> 1758</span>&#160;                                 <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01759"></a><span class="lineno"> 1759</span>&#160;        : TestAdditionQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l01760"></a><span class="lineno"> 1760</span>&#160;</div><div class="line"><a name="l01761"></a><span class="lineno"> 1761</span>&#160;        <span class="keywordtype">void</span> VisitConstantLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l01762"></a><span class="lineno"> 1762</span>&#160;                                <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a>&amp; input,</div><div class="line"><a name="l01763"></a><span class="lineno"> 1763</span>&#160;                                <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01764"></a><span class="lineno"> 1764</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01765"></a><span class="lineno"> 1765</span>&#160;            boost::ignore_unused(input, name);</div><div class="line"><a name="l01766"></a><span class="lineno"> 1766</span>&#160;            <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l01767"></a><span class="lineno"> 1767</span>&#160;</div><div class="line"><a name="l01768"></a><span class="lineno"> 1768</span>&#160;            <span class="comment">// Based off the range of values in the const tensor used for the test: [-2.0f, 6.0f]</span></div><div class="line"><a name="l01769"></a><span class="lineno"> 1769</span>&#160;            TestQuantizationParams(info, {8.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 64},</div><div class="line"><a name="l01770"></a><span class="lineno"> 1770</span>&#160;                                         {8.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>, -64},</div><div class="line"><a name="l01771"></a><span class="lineno"> 1771</span>&#160;                                         {6.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0},</div><div class="line"><a name="l01772"></a><span class="lineno"> 1772</span>&#160;                                         {6.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0});</div><div class="line"><a name="l01773"></a><span class="lineno"> 1773</span>&#160;        }</div><div class="line"><a name="l01774"></a><span class="lineno"> 1774</span>&#160;    };</div><div class="line"><a name="l01775"></a><span class="lineno"> 1775</span>&#160;</div><div class="line"><a name="l01776"></a><span class="lineno"> 1776</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l01777"></a><span class="lineno"> 1777</span>&#160;</div><div class="line"><a name="l01778"></a><span class="lineno"> 1778</span>&#160;    <span class="comment">// Constant layer data</span></div><div class="line"><a name="l01779"></a><span class="lineno"> 1779</span>&#160;    std::vector&lt;float&gt; data = {-2.0f, -1.0f, 0.0f, 1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f};</div><div class="line"><a name="l01780"></a><span class="lineno"> 1780</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U, 1U, 3U, 3U};</div><div class="line"><a name="l01781"></a><span class="lineno"> 1781</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> tensorInfo(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l01782"></a><span class="lineno"> 1782</span>&#160;    <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a> constantTensor(tensorInfo, data);</div><div class="line"><a name="l01783"></a><span class="lineno"> 1783</span>&#160;</div><div class="line"><a name="l01784"></a><span class="lineno"> 1784</span>&#160;    <span class="comment">// Add the layers</span></div><div class="line"><a name="l01785"></a><span class="lineno"> 1785</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input    = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l01786"></a><span class="lineno"> 1786</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* constant = network-&gt;AddConstantLayer(constantTensor);</div><div class="line"><a name="l01787"></a><span class="lineno"> 1787</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* addition = network-&gt;AddAdditionLayer();</div><div class="line"><a name="l01788"></a><span class="lineno"> 1788</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output   = network-&gt;AddOutputLayer(1);</div><div class="line"><a name="l01789"></a><span class="lineno"> 1789</span>&#160;</div><div class="line"><a name="l01790"></a><span class="lineno"> 1790</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l01791"></a><span class="lineno"> 1791</span>&#160;    input-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01792"></a><span class="lineno"> 1792</span>&#160;    constant-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(1));</div><div class="line"><a name="l01793"></a><span class="lineno"> 1793</span>&#160;    addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01794"></a><span class="lineno"> 1794</span>&#160;</div><div class="line"><a name="l01795"></a><span class="lineno"> 1795</span>&#160;    <span class="comment">// Set TensorInfo in the remaining layers</span></div><div class="line"><a name="l01796"></a><span class="lineno"> 1796</span>&#160;    input-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(tensorInfo);</div><div class="line"><a name="l01797"></a><span class="lineno"> 1797</span>&#160;    addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(tensorInfo);</div><div class="line"><a name="l01798"></a><span class="lineno"> 1798</span>&#160;    constant-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(tensorInfo);</div><div class="line"><a name="l01799"></a><span class="lineno"> 1799</span>&#160;</div><div class="line"><a name="l01800"></a><span class="lineno"> 1800</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l01801"></a><span class="lineno"> 1801</span>&#160;    TestConstantQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l01802"></a><span class="lineno"> 1802</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l01803"></a><span class="lineno"> 1803</span>&#160;</div><div class="line"><a name="l01804"></a><span class="lineno"> 1804</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l01805"></a><span class="lineno"> 1805</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01806"></a><span class="lineno"> 1806</span>&#160;    TestConstantQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l01807"></a><span class="lineno"> 1807</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l01808"></a><span class="lineno"> 1808</span>&#160;</div><div class="line"><a name="l01809"></a><span class="lineno"> 1809</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l01810"></a><span class="lineno"> 1810</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01811"></a><span class="lineno"> 1811</span>&#160;    TestConstantQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l01812"></a><span class="lineno"> 1812</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l01813"></a><span class="lineno"> 1813</span>&#160;</div><div class="line"><a name="l01814"></a><span class="lineno"> 1814</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l01815"></a><span class="lineno"> 1815</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l01816"></a><span class="lineno"> 1816</span>&#160;    TestConstantQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l01817"></a><span class="lineno"> 1817</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l01818"></a><span class="lineno"> 1818</span>&#160;}</div><div class="line"><a name="l01819"></a><span class="lineno"> 1819</span>&#160;</div><div class="line"><a name="l01820"></a><span class="lineno"><a class="line" href="namespacearmnn.html#ae3af95ea62252012cf93a98167afef64"> 1820</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeArgMinMax)</div><div class="line"><a name="l01821"></a><span class="lineno"> 1821</span>&#160;{</div><div class="line"><a name="l01822"></a><span class="lineno"> 1822</span>&#160;    <span class="keyword">class </span>TestArgMinMaxQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l01823"></a><span class="lineno"> 1823</span>&#160;    {</div><div class="line"><a name="l01824"></a><span class="lineno"> 1824</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l01825"></a><span class="lineno"> 1825</span>&#160;        TestArgMinMaxQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)  :</div><div class="line"><a name="l01826"></a><span class="lineno"> 1826</span>&#160;                TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l01827"></a><span class="lineno"> 1827</span>&#160;</div><div class="line"><a name="l01828"></a><span class="lineno"> 1828</span>&#160;        TestArgMinMaxQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l01829"></a><span class="lineno"> 1829</span>&#160;                                  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l01830"></a><span class="lineno"> 1830</span>&#160;                                  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape) :</div><div class="line"><a name="l01831"></a><span class="lineno"> 1831</span>&#160;                TestQuantization(options, inputShape, outputShape)</div><div class="line"><a name="l01832"></a><span class="lineno"> 1832</span>&#160;        {}</div><div class="line"><a name="l01833"></a><span class="lineno"> 1833</span>&#160;</div><div class="line"><a name="l01834"></a><span class="lineno"> 1834</span>&#160;        <span class="keywordtype">void</span> VisitInputLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l01835"></a><span class="lineno"> 1835</span>&#160;                             <a class="code" href="namespacearmnn.html#ab8cf8f9fb6792e654c2d8d8382f6f01b">LayerBindingId</a> <span class="keywordtype">id</span>,</div><div class="line"><a name="l01836"></a><span class="lineno"> 1836</span>&#160;                             <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01837"></a><span class="lineno"> 1837</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01838"></a><span class="lineno"> 1838</span>&#160;            boost::ignore_unused(layer, <span class="keywordtype">id</span>, name);</div><div class="line"><a name="l01839"></a><span class="lineno"> 1839</span>&#160;        }</div><div class="line"><a name="l01840"></a><span class="lineno"> 1840</span>&#160;</div><div class="line"><a name="l01841"></a><span class="lineno"> 1841</span>&#160;        <span class="keywordtype">void</span> VisitOutputLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l01842"></a><span class="lineno"> 1842</span>&#160;                              <a class="code" href="namespacearmnn.html#ab8cf8f9fb6792e654c2d8d8382f6f01b">LayerBindingId</a> <span class="keywordtype">id</span>,</div><div class="line"><a name="l01843"></a><span class="lineno"> 1843</span>&#160;                              <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01844"></a><span class="lineno"> 1844</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01845"></a><span class="lineno"> 1845</span>&#160;            boost::ignore_unused(layer, <span class="keywordtype">id</span>, name);</div><div class="line"><a name="l01846"></a><span class="lineno"> 1846</span>&#160;        }</div><div class="line"><a name="l01847"></a><span class="lineno"> 1847</span>&#160;        <span class="keywordtype">void</span> VisitArgMinMaxLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l01848"></a><span class="lineno"> 1848</span>&#160;                                 <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_arg_min_max_descriptor.html">ArgMinMaxDescriptor</a>&amp; argMinMaxDescriptor,</div><div class="line"><a name="l01849"></a><span class="lineno"> 1849</span>&#160;                                 <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01850"></a><span class="lineno"> 1850</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01851"></a><span class="lineno"> 1851</span>&#160;                boost::ignore_unused(argMinMaxDescriptor, name);</div><div class="line"><a name="l01852"></a><span class="lineno"> 1852</span>&#160;                <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> outputInfo = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l01853"></a><span class="lineno"> 1853</span>&#160;</div><div class="line"><a name="l01854"></a><span class="lineno"> 1854</span>&#160;                TestQuantizationParams(outputInfo,</div><div class="line"><a name="l01855"></a><span class="lineno"> 1855</span>&#160;                                       { 30.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 128 },</div><div class="line"><a name="l01856"></a><span class="lineno"> 1856</span>&#160;                                       { 30.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>,  0},</div><div class="line"><a name="l01857"></a><span class="lineno"> 1857</span>&#160;                                       { 15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0},</div><div class="line"><a name="l01858"></a><span class="lineno"> 1858</span>&#160;                                       { 15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0 });</div><div class="line"><a name="l01859"></a><span class="lineno"> 1859</span>&#160;        }</div><div class="line"><a name="l01860"></a><span class="lineno"> 1860</span>&#160;    };</div><div class="line"><a name="l01861"></a><span class="lineno"> 1861</span>&#160;</div><div class="line"><a name="l01862"></a><span class="lineno"> 1862</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l01863"></a><span class="lineno"> 1863</span>&#160;</div><div class="line"><a name="l01864"></a><span class="lineno"> 1864</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> inputShape{ 1, 1, 1, 5 };</div><div class="line"><a name="l01865"></a><span class="lineno"> 1865</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> outputShape{ 1, 1, 1 };</div><div class="line"><a name="l01866"></a><span class="lineno"> 1866</span>&#160;</div><div class="line"><a name="l01867"></a><span class="lineno"> 1867</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> inputInfo(inputShape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l01868"></a><span class="lineno"> 1868</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> outputInfo(outputShape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l01869"></a><span class="lineno"> 1869</span>&#160;</div><div class="line"><a name="l01870"></a><span class="lineno"> 1870</span>&#160;    <span class="comment">// Add the input layers</span></div><div class="line"><a name="l01871"></a><span class="lineno"> 1871</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l01872"></a><span class="lineno"> 1872</span>&#160;</div><div class="line"><a name="l01873"></a><span class="lineno"> 1873</span>&#160;    <span class="comment">// Add the layer under test</span></div><div class="line"><a name="l01874"></a><span class="lineno"> 1874</span>&#160;    <a class="code" href="structarmnn_1_1_arg_min_max_descriptor.html">ArgMinMaxDescriptor</a> argMinMaxDescriptor;</div><div class="line"><a name="l01875"></a><span class="lineno"> 1875</span>&#160;    argMinMaxDescriptor.<a class="code" href="structarmnn_1_1_arg_min_max_descriptor.html#ab1ae6f520bb1a4da191a0ae907477f23">m_Function</a> = <a class="code" href="namespacearmnn.html#ae7e8cbf71db6a490789ca6dcaa8deeaea6a061313d22e51e0f25b7cd4dc065233">ArgMinMaxFunction::Max</a>;</div><div class="line"><a name="l01876"></a><span class="lineno"> 1876</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* argMinMaxLayer = network-&gt;AddArgMinMaxLayer(argMinMaxDescriptor);</div><div class="line"><a name="l01877"></a><span class="lineno"> 1877</span>&#160;</div><div class="line"><a name="l01878"></a><span class="lineno"> 1878</span>&#160;    <span class="comment">// Add the output layers</span></div><div class="line"><a name="l01879"></a><span class="lineno"> 1879</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;AddOutputLayer(1);</div><div class="line"><a name="l01880"></a><span class="lineno"> 1880</span>&#160;</div><div class="line"><a name="l01881"></a><span class="lineno"> 1881</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l01882"></a><span class="lineno"> 1882</span>&#160;    input-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(argMinMaxLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01883"></a><span class="lineno"> 1883</span>&#160;    argMinMaxLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01884"></a><span class="lineno"> 1884</span>&#160;</div><div class="line"><a name="l01885"></a><span class="lineno"> 1885</span>&#160;    <span class="comment">// Set tensor info</span></div><div class="line"><a name="l01886"></a><span class="lineno"> 1886</span>&#160;    input-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(inputInfo);</div><div class="line"><a name="l01887"></a><span class="lineno"> 1887</span>&#160;    argMinMaxLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(outputInfo);</div><div class="line"><a name="l01888"></a><span class="lineno"> 1888</span>&#160;</div><div class="line"><a name="l01889"></a><span class="lineno"> 1889</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l01890"></a><span class="lineno"> 1890</span>&#160;    TestArgMinMaxQuantization validatorQAsymmU8(inputShape, outputShape);</div><div class="line"><a name="l01891"></a><span class="lineno"> 1891</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l01892"></a><span class="lineno"> 1892</span>&#160;</div><div class="line"><a name="l01893"></a><span class="lineno"> 1893</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l01894"></a><span class="lineno"> 1894</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01895"></a><span class="lineno"> 1895</span>&#160;    TestArgMinMaxQuantization validatorQAsymmS8(qAsymmS8Options, inputShape, outputShape);</div><div class="line"><a name="l01896"></a><span class="lineno"> 1896</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l01897"></a><span class="lineno"> 1897</span>&#160;</div><div class="line"><a name="l01898"></a><span class="lineno"> 1898</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l01899"></a><span class="lineno"> 1899</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01900"></a><span class="lineno"> 1900</span>&#160;    TestArgMinMaxQuantization validatorQSymmS8(qSymmS8Options, inputShape, outputShape);</div><div class="line"><a name="l01901"></a><span class="lineno"> 1901</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l01902"></a><span class="lineno"> 1902</span>&#160;</div><div class="line"><a name="l01903"></a><span class="lineno"> 1903</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l01904"></a><span class="lineno"> 1904</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l01905"></a><span class="lineno"> 1905</span>&#160;    TestArgMinMaxQuantization validatorQSymmS16(qSymmS16options, inputShape, outputShape);</div><div class="line"><a name="l01906"></a><span class="lineno"> 1906</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l01907"></a><span class="lineno"> 1907</span>&#160;}</div><div class="line"><a name="l01908"></a><span class="lineno"> 1908</span>&#160;</div><div class="line"><a name="l01909"></a><span class="lineno"><a class="line" href="namespacearmnn.html#ab83f837cdd5bfcff537dae72a96d6fc8"> 1909</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeComparison)</div><div class="line"><a name="l01910"></a><span class="lineno"> 1910</span>&#160;{</div><div class="line"><a name="l01911"></a><span class="lineno"> 1911</span>&#160;    <span class="keyword">class </span>TestComparisonQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l01912"></a><span class="lineno"> 1912</span>&#160;    {</div><div class="line"><a name="l01913"></a><span class="lineno"> 1913</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l01914"></a><span class="lineno"> 1914</span>&#160;        TestComparisonQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01915"></a><span class="lineno"> 1915</span>&#160;            : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l01916"></a><span class="lineno"> 1916</span>&#160;</div><div class="line"><a name="l01917"></a><span class="lineno"> 1917</span>&#160;        TestComparisonQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l01918"></a><span class="lineno"> 1918</span>&#160;                                   <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l01919"></a><span class="lineno"> 1919</span>&#160;                                   <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01920"></a><span class="lineno"> 1920</span>&#160;            : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l01921"></a><span class="lineno"> 1921</span>&#160;</div><div class="line"><a name="l01922"></a><span class="lineno"> 1922</span>&#160;        <span class="keywordtype">void</span> VisitComparisonLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l01923"></a><span class="lineno"> 1923</span>&#160;                                  <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_comparison_descriptor.html">ComparisonDescriptor</a>&amp; descriptor,</div><div class="line"><a name="l01924"></a><span class="lineno"> 1924</span>&#160;                                  <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01925"></a><span class="lineno"> 1925</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01926"></a><span class="lineno"> 1926</span>&#160;            boost::ignore_unused(descriptor, name);</div><div class="line"><a name="l01927"></a><span class="lineno"> 1927</span>&#160;            <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l01928"></a><span class="lineno"> 1928</span>&#160;</div><div class="line"><a name="l01929"></a><span class="lineno"> 1929</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qAsymmU8Params{ 30.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 128 };</div><div class="line"><a name="l01930"></a><span class="lineno"> 1930</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qAsymmS8Params { 30.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>,  0};</div><div class="line"><a name="l01931"></a><span class="lineno"> 1931</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qSymmS8Params { 15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0};</div><div class="line"><a name="l01932"></a><span class="lineno"> 1932</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qSymmS16Params{ 15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0 };</div><div class="line"><a name="l01933"></a><span class="lineno"> 1933</span>&#160;</div><div class="line"><a name="l01934"></a><span class="lineno"> 1934</span>&#160;            TestQuantizationParams(info, qAsymmU8Params, qAsymmS8Params, qSymmS8Params, qSymmS16Params);</div><div class="line"><a name="l01935"></a><span class="lineno"> 1935</span>&#160;        }</div><div class="line"><a name="l01936"></a><span class="lineno"> 1936</span>&#160;    };</div><div class="line"><a name="l01937"></a><span class="lineno"> 1937</span>&#160;</div><div class="line"><a name="l01938"></a><span class="lineno"> 1938</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> tensorShape{ 1u };</div><div class="line"><a name="l01939"></a><span class="lineno"> 1939</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> tensorInfo(tensorShape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l01940"></a><span class="lineno"> 1940</span>&#160;</div><div class="line"><a name="l01941"></a><span class="lineno"> 1941</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l01942"></a><span class="lineno"> 1942</span>&#160;    <a class="code" href="structarmnn_1_1_comparison_descriptor.html">ComparisonDescriptor</a> descriptor(<a class="code" href="namespacearmnn.html#a2d299363c9fc33334c571fa29ca4f58caa4cbdbb6070a5abb35fc95ecf1e22c14">ComparisonOperation::LessOrEqual</a>);</div><div class="line"><a name="l01943"></a><span class="lineno"> 1943</span>&#160;</div><div class="line"><a name="l01944"></a><span class="lineno"> 1944</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* inputLayer0     = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l01945"></a><span class="lineno"> 1945</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* inputLayer1     = network-&gt;AddInputLayer(1);</div><div class="line"><a name="l01946"></a><span class="lineno"> 1946</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* comparisonLayer = network-&gt;AddComparisonLayer(descriptor);</div><div class="line"><a name="l01947"></a><span class="lineno"> 1947</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* outputLayer     = network-&gt;AddOutputLayer(0);</div><div class="line"><a name="l01948"></a><span class="lineno"> 1948</span>&#160;</div><div class="line"><a name="l01949"></a><span class="lineno"> 1949</span>&#160;    inputLayer0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(comparisonLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01950"></a><span class="lineno"> 1950</span>&#160;    inputLayer1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(comparisonLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(1));</div><div class="line"><a name="l01951"></a><span class="lineno"> 1951</span>&#160;    comparisonLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(outputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l01952"></a><span class="lineno"> 1952</span>&#160;</div><div class="line"><a name="l01953"></a><span class="lineno"> 1953</span>&#160;    inputLayer0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(tensorInfo);</div><div class="line"><a name="l01954"></a><span class="lineno"> 1954</span>&#160;    inputLayer1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(tensorInfo);</div><div class="line"><a name="l01955"></a><span class="lineno"> 1955</span>&#160;    comparisonLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(tensorInfo);</div><div class="line"><a name="l01956"></a><span class="lineno"> 1956</span>&#160;</div><div class="line"><a name="l01957"></a><span class="lineno"> 1957</span>&#160;    <span class="comment">// test QAsymmU8 quantization</span></div><div class="line"><a name="l01958"></a><span class="lineno"> 1958</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l01959"></a><span class="lineno"> 1959</span>&#160;    TestComparisonQuantization validatorQAsymmU8(tensorShape, tensorShape);</div><div class="line"><a name="l01960"></a><span class="lineno"> 1960</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l01961"></a><span class="lineno"> 1961</span>&#160;</div><div class="line"><a name="l01962"></a><span class="lineno"> 1962</span>&#160;    <span class="comment">// test QAsymmS8 quantization</span></div><div class="line"><a name="l01963"></a><span class="lineno"> 1963</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l01964"></a><span class="lineno"> 1964</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01965"></a><span class="lineno"> 1965</span>&#160;    TestComparisonQuantization validatorQAsymmS8(qAsymmS8Options, tensorShape, tensorShape);</div><div class="line"><a name="l01966"></a><span class="lineno"> 1966</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l01967"></a><span class="lineno"> 1967</span>&#160;</div><div class="line"><a name="l01968"></a><span class="lineno"> 1968</span>&#160;    <span class="comment">// test QSymmS8 quantization</span></div><div class="line"><a name="l01969"></a><span class="lineno"> 1969</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l01970"></a><span class="lineno"> 1970</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l01971"></a><span class="lineno"> 1971</span>&#160;    TestComparisonQuantization validatorQSymmS8(qSymmS8Options, tensorShape, tensorShape);</div><div class="line"><a name="l01972"></a><span class="lineno"> 1972</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l01973"></a><span class="lineno"> 1973</span>&#160;</div><div class="line"><a name="l01974"></a><span class="lineno"> 1974</span>&#160;    <span class="comment">// test QuantisedSymmS16 quantization</span></div><div class="line"><a name="l01975"></a><span class="lineno"> 1975</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l01976"></a><span class="lineno"> 1976</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l01977"></a><span class="lineno"> 1977</span>&#160;    TestComparisonQuantization validatorQSymmS16(qSymmS16options, tensorShape, tensorShape);</div><div class="line"><a name="l01978"></a><span class="lineno"> 1978</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l01979"></a><span class="lineno"> 1979</span>&#160;}</div><div class="line"><a name="l01980"></a><span class="lineno"> 1980</span>&#160;</div><div class="line"><a name="l01981"></a><span class="lineno"><a class="line" href="namespacearmnn.html#add47ebcd4a59304a25c71996aea2b38b"> 1981</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeConcat)</div><div class="line"><a name="l01982"></a><span class="lineno"> 1982</span>&#160;{</div><div class="line"><a name="l01983"></a><span class="lineno"> 1983</span>&#160;    <span class="keyword">class </span>TestConcatQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l01984"></a><span class="lineno"> 1984</span>&#160;    {</div><div class="line"><a name="l01985"></a><span class="lineno"> 1985</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l01986"></a><span class="lineno"> 1986</span>&#160;        TestConcatQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01987"></a><span class="lineno"> 1987</span>&#160;        : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l01988"></a><span class="lineno"> 1988</span>&#160;</div><div class="line"><a name="l01989"></a><span class="lineno"> 1989</span>&#160;        TestConcatQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l01990"></a><span class="lineno"> 1990</span>&#160;                               <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l01991"></a><span class="lineno"> 1991</span>&#160;                               <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l01992"></a><span class="lineno"> 1992</span>&#160;        : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l01993"></a><span class="lineno"> 1993</span>&#160;</div><div class="line"><a name="l01994"></a><span class="lineno"> 1994</span>&#160;        <span class="keywordtype">void</span> VisitInputLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l01995"></a><span class="lineno"> 1995</span>&#160;                             <a class="code" href="namespacearmnn.html#ab8cf8f9fb6792e654c2d8d8382f6f01b">LayerBindingId</a> <span class="keywordtype">id</span>,</div><div class="line"><a name="l01996"></a><span class="lineno"> 1996</span>&#160;                             <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l01997"></a><span class="lineno"> 1997</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l01998"></a><span class="lineno"> 1998</span>&#160;            boost::ignore_unused(layer, <span class="keywordtype">id</span>, name);</div><div class="line"><a name="l01999"></a><span class="lineno"> 1999</span>&#160;        }</div><div class="line"><a name="l02000"></a><span class="lineno"> 2000</span>&#160;        <span class="keywordtype">void</span> VisitOutputLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02001"></a><span class="lineno"> 2001</span>&#160;                              <a class="code" href="namespacearmnn.html#ab8cf8f9fb6792e654c2d8d8382f6f01b">LayerBindingId</a> <span class="keywordtype">id</span>,</div><div class="line"><a name="l02002"></a><span class="lineno"> 2002</span>&#160;                              <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02003"></a><span class="lineno"> 2003</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l02004"></a><span class="lineno"> 2004</span>&#160;            boost::ignore_unused(layer, <span class="keywordtype">id</span>, name);</div><div class="line"><a name="l02005"></a><span class="lineno"> 2005</span>&#160;        }</div><div class="line"><a name="l02006"></a><span class="lineno"> 2006</span>&#160;        <span class="keywordtype">void</span> VisitConcatLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02007"></a><span class="lineno"> 2007</span>&#160;                              <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_origins_descriptor.html">OriginsDescriptor</a>&amp; originsDescriptor,</div><div class="line"><a name="l02008"></a><span class="lineno"> 2008</span>&#160;                              <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02009"></a><span class="lineno"> 2009</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l02010"></a><span class="lineno"> 2010</span>&#160;            boost::ignore_unused(originsDescriptor, name);</div><div class="line"><a name="l02011"></a><span class="lineno"> 2011</span>&#160;            <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> outputInfo = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l02012"></a><span class="lineno"> 2012</span>&#160;            TestQuantizationParams(</div><div class="line"><a name="l02013"></a><span class="lineno"> 2013</span>&#160;                outputInfo, {60.8f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 65},</div><div class="line"><a name="l02014"></a><span class="lineno"> 2014</span>&#160;                            {60.8f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  -63},</div><div class="line"><a name="l02015"></a><span class="lineno"> 2015</span>&#160;                            {45.3f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0},</div><div class="line"><a name="l02016"></a><span class="lineno"> 2016</span>&#160;                            {45.3f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0});</div><div class="line"><a name="l02017"></a><span class="lineno"> 2017</span>&#160;</div><div class="line"><a name="l02018"></a><span class="lineno"> 2018</span>&#160;            <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> inputInfo0 = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_input_slot.html#a81fbf6103761e55061b62ba989b00f10">GetConnection</a>()-&gt;<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l02019"></a><span class="lineno"> 2019</span>&#160;            <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> inputInfo1 = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(1).<a class="code" href="classarmnn_1_1_i_input_slot.html#a81fbf6103761e55061b62ba989b00f10">GetConnection</a>()-&gt;<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l02020"></a><span class="lineno"> 2020</span>&#160;            <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> inputInfo2 = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(2).<a class="code" href="classarmnn_1_1_i_input_slot.html#a81fbf6103761e55061b62ba989b00f10">GetConnection</a>()-&gt;<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l02021"></a><span class="lineno"> 2021</span>&#160;</div><div class="line"><a name="l02022"></a><span class="lineno"> 2022</span>&#160;            TestDifferentQuantizationScale(inputInfo0, inputInfo1);</div><div class="line"><a name="l02023"></a><span class="lineno"> 2023</span>&#160;            TestDifferentQuantizationScale(inputInfo0, inputInfo2);</div><div class="line"><a name="l02024"></a><span class="lineno"> 2024</span>&#160;            TestDifferentQuantizationScale(inputInfo1, inputInfo2);</div><div class="line"><a name="l02025"></a><span class="lineno"> 2025</span>&#160;            TestDifferentQuantizationScale(inputInfo0, outputInfo);</div><div class="line"><a name="l02026"></a><span class="lineno"> 2026</span>&#160;        }</div><div class="line"><a name="l02027"></a><span class="lineno"> 2027</span>&#160;    };</div><div class="line"><a name="l02028"></a><span class="lineno"> 2028</span>&#160;</div><div class="line"><a name="l02029"></a><span class="lineno"> 2029</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l02030"></a><span class="lineno"> 2030</span>&#160;</div><div class="line"><a name="l02031"></a><span class="lineno"> 2031</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input0 = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l02032"></a><span class="lineno"> 2032</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input1 = network-&gt;AddInputLayer(1);</div><div class="line"><a name="l02033"></a><span class="lineno"> 2033</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input2 = network-&gt;AddInputLayer(2);</div><div class="line"><a name="l02034"></a><span class="lineno"> 2034</span>&#160;</div><div class="line"><a name="l02035"></a><span class="lineno"> 2035</span>&#160;    <a class="code" href="structarmnn_1_1_origins_descriptor.html">OriginsDescriptor</a> descriptor(3, 1);</div><div class="line"><a name="l02036"></a><span class="lineno"> 2036</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* concatLayer = network-&gt;AddConcatLayer(descriptor);</div><div class="line"><a name="l02037"></a><span class="lineno"> 2037</span>&#160;</div><div class="line"><a name="l02038"></a><span class="lineno"> 2038</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output0 = network-&gt;AddOutputLayer(3);</div><div class="line"><a name="l02039"></a><span class="lineno"> 2039</span>&#160;</div><div class="line"><a name="l02040"></a><span class="lineno"> 2040</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l02041"></a><span class="lineno"> 2041</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(concatLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02042"></a><span class="lineno"> 2042</span>&#160;    input1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(concatLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(1));</div><div class="line"><a name="l02043"></a><span class="lineno"> 2043</span>&#160;    input2-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(concatLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(2));</div><div class="line"><a name="l02044"></a><span class="lineno"> 2044</span>&#160;    concatLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(output0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02045"></a><span class="lineno"> 2045</span>&#160;</div><div class="line"><a name="l02046"></a><span class="lineno"> 2046</span>&#160;    <span class="comment">// Set TensorInfo</span></div><div class="line"><a name="l02047"></a><span class="lineno"> 2047</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l02048"></a><span class="lineno"> 2048</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l02049"></a><span class="lineno"> 2049</span>&#160;</div><div class="line"><a name="l02050"></a><span class="lineno"> 2050</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l02051"></a><span class="lineno"> 2051</span>&#160;    input1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l02052"></a><span class="lineno"> 2052</span>&#160;    input2-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l02053"></a><span class="lineno"> 2053</span>&#160;    concatLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l02054"></a><span class="lineno"> 2054</span>&#160;</div><div class="line"><a name="l02055"></a><span class="lineno"> 2055</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l02056"></a><span class="lineno"> 2056</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l02057"></a><span class="lineno"> 2057</span>&#160;    <a class="code" href="namespacearmnn.html#a41119e261eec9343888d2ceab1e4999a">INetworkQuantizerPtr</a> quantizerPtrQAsymmU8 =  <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get());</div><div class="line"><a name="l02058"></a><span class="lineno"> 2058</span>&#160;    <a class="code" href="namespacearmnn.html#a41119e261eec9343888d2ceab1e4999a">INetworkQuantizerPtr</a> quantizerPtrQSymmS8  =  <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options);</div><div class="line"><a name="l02059"></a><span class="lineno"> 2059</span>&#160;    <a class="code" href="namespacearmnn.html#a41119e261eec9343888d2ceab1e4999a">INetworkQuantizerPtr</a> quantizerPtrQSymmS16 =  <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options);</div><div class="line"><a name="l02060"></a><span class="lineno"> 2060</span>&#160;    <span class="comment">// Override the input ranges</span></div><div class="line"><a name="l02061"></a><span class="lineno"> 2061</span>&#160;    <span class="keywordtype">float</span> min = -15.5f;</div><div class="line"><a name="l02062"></a><span class="lineno"> 2062</span>&#160;    <span class="keywordtype">float</span> max = 45.3f;</div><div class="line"><a name="l02063"></a><span class="lineno"> 2063</span>&#160;</div><div class="line"><a name="l02064"></a><span class="lineno"> 2064</span>&#160;    quantizerPtrQAsymmU8-&gt;OverrideInputRange(0, (min + 2.1f), (max - 3.2f));</div><div class="line"><a name="l02065"></a><span class="lineno"> 2065</span>&#160;    quantizerPtrQAsymmU8-&gt;OverrideInputRange(1, (min + 6.7f), max);</div><div class="line"><a name="l02066"></a><span class="lineno"> 2066</span>&#160;    quantizerPtrQAsymmU8-&gt;OverrideInputRange(2, min, (max - 7.8f));</div><div class="line"><a name="l02067"></a><span class="lineno"> 2067</span>&#160;</div><div class="line"><a name="l02068"></a><span class="lineno"> 2068</span>&#160;    quantizerPtrQSymmS8-&gt;OverrideInputRange(0, (min + 2.1f), (max - 3.2f));</div><div class="line"><a name="l02069"></a><span class="lineno"> 2069</span>&#160;    quantizerPtrQSymmS8-&gt;OverrideInputRange(1, (min + 6.7f), max);</div><div class="line"><a name="l02070"></a><span class="lineno"> 2070</span>&#160;    quantizerPtrQSymmS8-&gt;OverrideInputRange(2, min, (max - 7.8f));</div><div class="line"><a name="l02071"></a><span class="lineno"> 2071</span>&#160;</div><div class="line"><a name="l02072"></a><span class="lineno"> 2072</span>&#160;    quantizerPtrQSymmS16-&gt;OverrideInputRange(0, (min + 2.1f), (max - 3.2f));</div><div class="line"><a name="l02073"></a><span class="lineno"> 2073</span>&#160;    quantizerPtrQSymmS16-&gt;OverrideInputRange(1, (min + 6.7f), max);</div><div class="line"><a name="l02074"></a><span class="lineno"> 2074</span>&#160;    quantizerPtrQSymmS16-&gt;OverrideInputRange(2, min, (max - 7.8f));</div><div class="line"><a name="l02075"></a><span class="lineno"> 2075</span>&#160;</div><div class="line"><a name="l02076"></a><span class="lineno"> 2076</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = quantizerPtrQAsymmU8-&gt;ExportNetwork();</div><div class="line"><a name="l02077"></a><span class="lineno"> 2077</span>&#160;    TestConcatQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l02078"></a><span class="lineno"> 2078</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l02079"></a><span class="lineno"> 2079</span>&#160;</div><div class="line"><a name="l02080"></a><span class="lineno"> 2080</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = quantizerPtrQSymmS8-&gt;ExportNetwork();</div><div class="line"><a name="l02081"></a><span class="lineno"> 2081</span>&#160;    TestConcatQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l02082"></a><span class="lineno"> 2082</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l02083"></a><span class="lineno"> 2083</span>&#160;</div><div class="line"><a name="l02084"></a><span class="lineno"> 2084</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = quantizerPtrQSymmS16-&gt;ExportNetwork();</div><div class="line"><a name="l02085"></a><span class="lineno"> 2085</span>&#160;    TestConcatQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l02086"></a><span class="lineno"> 2086</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l02087"></a><span class="lineno"> 2087</span>&#160;}</div><div class="line"><a name="l02088"></a><span class="lineno"> 2088</span>&#160;</div><div class="line"><a name="l02089"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a9258afcd4c6d8443c9130d8c9bf26442"> 2089</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeReshape)</div><div class="line"><a name="l02090"></a><span class="lineno"> 2090</span>&#160;{</div><div class="line"><a name="l02091"></a><span class="lineno"> 2091</span>&#160;    <span class="keyword">class </span>TestReshapeQuantization : <span class="keyword">public</span> TestLeakyReLuActivationQuantization</div><div class="line"><a name="l02092"></a><span class="lineno"> 2092</span>&#160;    {</div><div class="line"><a name="l02093"></a><span class="lineno"> 2093</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l02094"></a><span class="lineno"> 2094</span>&#160;        TestReshapeQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02095"></a><span class="lineno"> 2095</span>&#160;        : TestLeakyReLuActivationQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l02096"></a><span class="lineno"> 2096</span>&#160;</div><div class="line"><a name="l02097"></a><span class="lineno"> 2097</span>&#160;        TestReshapeQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l02098"></a><span class="lineno"> 2098</span>&#160;                                <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l02099"></a><span class="lineno"> 2099</span>&#160;                                <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02100"></a><span class="lineno"> 2100</span>&#160;        : TestLeakyReLuActivationQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l02101"></a><span class="lineno"> 2101</span>&#160;</div><div class="line"><a name="l02102"></a><span class="lineno"> 2102</span>&#160;        <span class="keyword">virtual</span> <span class="keywordtype">void</span> VisitReshapeLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02103"></a><span class="lineno"> 2103</span>&#160;                                       <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_reshape_descriptor.html">ReshapeDescriptor</a>&amp; reshapeDescriptor,</div><div class="line"><a name="l02104"></a><span class="lineno"> 2104</span>&#160;                                       <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02105"></a><span class="lineno"> 2105</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l02106"></a><span class="lineno"> 2106</span>&#160;            boost::ignore_unused(reshapeDescriptor, name);</div><div class="line"><a name="l02107"></a><span class="lineno"> 2107</span>&#160;            CheckForwardedQuantizationSettings(layer);</div><div class="line"><a name="l02108"></a><span class="lineno"> 2108</span>&#160;        }</div><div class="line"><a name="l02109"></a><span class="lineno"> 2109</span>&#160;    };</div><div class="line"><a name="l02110"></a><span class="lineno"> 2110</span>&#160;</div><div class="line"><a name="l02111"></a><span class="lineno"> 2111</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l02112"></a><span class="lineno"> 2112</span>&#160;</div><div class="line"><a name="l02113"></a><span class="lineno"> 2113</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l02114"></a><span class="lineno"> 2114</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l02115"></a><span class="lineno"> 2115</span>&#160;</div><div class="line"><a name="l02116"></a><span class="lineno"> 2116</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* activation = <a class="code" href="namespacearmnn.html#a120c131df35d78b3a56cb0f07decaf35">CreateStartOfLeakyReluNetwork</a>(network.get(), <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l02117"></a><span class="lineno"> 2117</span>&#160;</div><div class="line"><a name="l02118"></a><span class="lineno"> 2118</span>&#160;    <span class="comment">// Add the layer under test</span></div><div class="line"><a name="l02119"></a><span class="lineno"> 2119</span>&#160;    <a class="code" href="structarmnn_1_1_reshape_descriptor.html">ReshapeDescriptor</a> descriptor({1, 2, 3, 4});</div><div class="line"><a name="l02120"></a><span class="lineno"> 2120</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* reshape = network-&gt;AddReshapeLayer(descriptor);</div><div class="line"><a name="l02121"></a><span class="lineno"> 2121</span>&#160;</div><div class="line"><a name="l02122"></a><span class="lineno"> 2122</span>&#160;    <a class="code" href="namespacearmnn.html#a6fff4b4b1b5d4d37c9cf53d0e31c05dd">CompleteLeakyReluNetwork</a>(network.get(), activation, reshape, <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l02123"></a><span class="lineno"> 2123</span>&#160;</div><div class="line"><a name="l02124"></a><span class="lineno"> 2124</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l02125"></a><span class="lineno"> 2125</span>&#160;    TestReshapeQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l02126"></a><span class="lineno"> 2126</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l02127"></a><span class="lineno"> 2127</span>&#160;</div><div class="line"><a name="l02128"></a><span class="lineno"> 2128</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l02129"></a><span class="lineno"> 2129</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02130"></a><span class="lineno"> 2130</span>&#160;    TestReshapeQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l02131"></a><span class="lineno"> 2131</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l02132"></a><span class="lineno"> 2132</span>&#160;</div><div class="line"><a name="l02133"></a><span class="lineno"> 2133</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l02134"></a><span class="lineno"> 2134</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02135"></a><span class="lineno"> 2135</span>&#160;    TestReshapeQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l02136"></a><span class="lineno"> 2136</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l02137"></a><span class="lineno"> 2137</span>&#160;</div><div class="line"><a name="l02138"></a><span class="lineno"> 2138</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l02139"></a><span class="lineno"> 2139</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l02140"></a><span class="lineno"> 2140</span>&#160;    TestReshapeQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l02141"></a><span class="lineno"> 2141</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l02142"></a><span class="lineno"> 2142</span>&#160;}</div><div class="line"><a name="l02143"></a><span class="lineno"> 2143</span>&#160;</div><div class="line"><a name="l02144"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a23a4f3c387a2a3a035e97764e34277c6"> 2144</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeSplitter)</div><div class="line"><a name="l02145"></a><span class="lineno"> 2145</span>&#160;{</div><div class="line"><a name="l02146"></a><span class="lineno"> 2146</span>&#160;    <span class="keyword">class </span>TestSplitterQuantization : <span class="keyword">public</span> TestLeakyReLuActivationQuantization</div><div class="line"><a name="l02147"></a><span class="lineno"> 2147</span>&#160;    {</div><div class="line"><a name="l02148"></a><span class="lineno"> 2148</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l02149"></a><span class="lineno"> 2149</span>&#160;        TestSplitterQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02150"></a><span class="lineno"> 2150</span>&#160;        : TestLeakyReLuActivationQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l02151"></a><span class="lineno"> 2151</span>&#160;</div><div class="line"><a name="l02152"></a><span class="lineno"> 2152</span>&#160;        TestSplitterQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l02153"></a><span class="lineno"> 2153</span>&#160;                                 <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l02154"></a><span class="lineno"> 2154</span>&#160;                                 <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02155"></a><span class="lineno"> 2155</span>&#160;        : TestLeakyReLuActivationQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l02156"></a><span class="lineno"> 2156</span>&#160;</div><div class="line"><a name="l02157"></a><span class="lineno"> 2157</span>&#160;        <span class="keyword">virtual</span> <span class="keywordtype">void</span> VisitSplitterLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02158"></a><span class="lineno"> 2158</span>&#160;                                        <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_views_descriptor.html">SplitterDescriptor</a>&amp; desc,</div><div class="line"><a name="l02159"></a><span class="lineno"> 2159</span>&#160;                                        <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)</div><div class="line"><a name="l02160"></a><span class="lineno"> 2160</span>&#160;        {</div><div class="line"><a name="l02161"></a><span class="lineno"> 2161</span>&#160;            boost::ignore_unused(desc, name);</div><div class="line"><a name="l02162"></a><span class="lineno"> 2162</span>&#160;            CheckForwardedQuantizationSettings(layer);</div><div class="line"><a name="l02163"></a><span class="lineno"> 2163</span>&#160;        }</div><div class="line"><a name="l02164"></a><span class="lineno"> 2164</span>&#160;    };</div><div class="line"><a name="l02165"></a><span class="lineno"> 2165</span>&#160;</div><div class="line"><a name="l02166"></a><span class="lineno"> 2166</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l02167"></a><span class="lineno"> 2167</span>&#160;</div><div class="line"><a name="l02168"></a><span class="lineno"> 2168</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{3U};</div><div class="line"><a name="l02169"></a><span class="lineno"> 2169</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l02170"></a><span class="lineno"> 2170</span>&#160;</div><div class="line"><a name="l02171"></a><span class="lineno"> 2171</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* activation = <a class="code" href="namespacearmnn.html#a120c131df35d78b3a56cb0f07decaf35">CreateStartOfLeakyReluNetwork</a>(network.get(), <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l02172"></a><span class="lineno"> 2172</span>&#160;</div><div class="line"><a name="l02173"></a><span class="lineno"> 2173</span>&#160;    <span class="comment">// Add the layer under test</span></div><div class="line"><a name="l02174"></a><span class="lineno"> 2174</span>&#160;    <a class="code" href="structarmnn_1_1_views_descriptor.html">ViewsDescriptor</a> splitterDesc(2,4);</div><div class="line"><a name="l02175"></a><span class="lineno"> 2175</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* splitter = network-&gt;AddSplitterLayer(splitterDesc);</div><div class="line"><a name="l02176"></a><span class="lineno"> 2176</span>&#160;    <a class="code" href="namespacearmnn.html#a6fff4b4b1b5d4d37c9cf53d0e31c05dd">CompleteLeakyReluNetwork</a>(network.get(), activation, splitter, <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l02177"></a><span class="lineno"> 2177</span>&#160;</div><div class="line"><a name="l02178"></a><span class="lineno"> 2178</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l02179"></a><span class="lineno"> 2179</span>&#160;    TestSplitterQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l02180"></a><span class="lineno"> 2180</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l02181"></a><span class="lineno"> 2181</span>&#160;</div><div class="line"><a name="l02182"></a><span class="lineno"> 2182</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l02183"></a><span class="lineno"> 2183</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02184"></a><span class="lineno"> 2184</span>&#160;    TestSplitterQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l02185"></a><span class="lineno"> 2185</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l02186"></a><span class="lineno"> 2186</span>&#160;</div><div class="line"><a name="l02187"></a><span class="lineno"> 2187</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l02188"></a><span class="lineno"> 2188</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02189"></a><span class="lineno"> 2189</span>&#160;    TestSplitterQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l02190"></a><span class="lineno"> 2190</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l02191"></a><span class="lineno"> 2191</span>&#160;</div><div class="line"><a name="l02192"></a><span class="lineno"> 2192</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l02193"></a><span class="lineno"> 2193</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l02194"></a><span class="lineno"> 2194</span>&#160;    TestSplitterQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l02195"></a><span class="lineno"> 2195</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l02196"></a><span class="lineno"> 2196</span>&#160;}</div><div class="line"><a name="l02197"></a><span class="lineno"> 2197</span>&#160;</div><div class="line"><a name="l02198"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a102f37a09de1b0d4d78740a3c12902bf"> 2198</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeResize)</div><div class="line"><a name="l02199"></a><span class="lineno"> 2199</span>&#160;{</div><div class="line"><a name="l02200"></a><span class="lineno"> 2200</span>&#160;    <span class="keyword">class </span>TestResizeQuantization : <span class="keyword">public</span> TestLeakyReLuActivationQuantization</div><div class="line"><a name="l02201"></a><span class="lineno"> 2201</span>&#160;    {</div><div class="line"><a name="l02202"></a><span class="lineno"> 2202</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l02203"></a><span class="lineno"> 2203</span>&#160;        TestResizeQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02204"></a><span class="lineno"> 2204</span>&#160;                : TestLeakyReLuActivationQuantization(inputShape, outputShape)</div><div class="line"><a name="l02205"></a><span class="lineno"> 2205</span>&#160;        {}</div><div class="line"><a name="l02206"></a><span class="lineno"> 2206</span>&#160;</div><div class="line"><a name="l02207"></a><span class="lineno"> 2207</span>&#160;        TestResizeQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l02208"></a><span class="lineno"> 2208</span>&#160;                                       <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l02209"></a><span class="lineno"> 2209</span>&#160;                                       <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02210"></a><span class="lineno"> 2210</span>&#160;                : TestLeakyReLuActivationQuantization(options, inputShape, outputShape)</div><div class="line"><a name="l02211"></a><span class="lineno"> 2211</span>&#160;        {}</div><div class="line"><a name="l02212"></a><span class="lineno"> 2212</span>&#160;</div><div class="line"><a name="l02213"></a><span class="lineno"> 2213</span>&#160;        <span class="keywordtype">void</span> VisitResizeLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02214"></a><span class="lineno"> 2214</span>&#160;                                      <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_resize_descriptor.html">ResizeDescriptor</a>&amp; resizeDescriptor,</div><div class="line"><a name="l02215"></a><span class="lineno"> 2215</span>&#160;                                      <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02216"></a><span class="lineno"> 2216</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l02217"></a><span class="lineno"> 2217</span>&#160;            boost::ignore_unused(resizeDescriptor, name);</div><div class="line"><a name="l02218"></a><span class="lineno"> 2218</span>&#160;            CheckForwardedQuantizationSettings(layer);</div><div class="line"><a name="l02219"></a><span class="lineno"> 2219</span>&#160;        }</div><div class="line"><a name="l02220"></a><span class="lineno"> 2220</span>&#160;    };</div><div class="line"><a name="l02221"></a><span class="lineno"> 2221</span>&#160;</div><div class="line"><a name="l02222"></a><span class="lineno"> 2222</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l02223"></a><span class="lineno"> 2223</span>&#160;</div><div class="line"><a name="l02224"></a><span class="lineno"> 2224</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l02225"></a><span class="lineno"> 2225</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l02226"></a><span class="lineno"> 2226</span>&#160;</div><div class="line"><a name="l02227"></a><span class="lineno"> 2227</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* activation = <a class="code" href="namespacearmnn.html#a120c131df35d78b3a56cb0f07decaf35">CreateStartOfLeakyReluNetwork</a>(network.get(), <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l02228"></a><span class="lineno"> 2228</span>&#160;</div><div class="line"><a name="l02229"></a><span class="lineno"> 2229</span>&#160;    <span class="comment">// Add the layer under test</span></div><div class="line"><a name="l02230"></a><span class="lineno"> 2230</span>&#160;    <a class="code" href="structarmnn_1_1_resize_descriptor.html">ResizeDescriptor</a> descriptor;</div><div class="line"><a name="l02231"></a><span class="lineno"> 2231</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_resize_descriptor.html#a46c3fa15c46fb0d1dcdc24d0ea5cb5cd">m_TargetHeight</a> = 3;</div><div class="line"><a name="l02232"></a><span class="lineno"> 2232</span>&#160;    descriptor.m_TargetWidth  = 3;</div><div class="line"><a name="l02233"></a><span class="lineno"> 2233</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* resizeLayer = network-&gt;AddResizeLayer(descriptor);</div><div class="line"><a name="l02234"></a><span class="lineno"> 2234</span>&#160;</div><div class="line"><a name="l02235"></a><span class="lineno"> 2235</span>&#160;    <a class="code" href="namespacearmnn.html#a6fff4b4b1b5d4d37c9cf53d0e31c05dd">CompleteLeakyReluNetwork</a>(network.get(), activation, resizeLayer, <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l02236"></a><span class="lineno"> 2236</span>&#160;</div><div class="line"><a name="l02237"></a><span class="lineno"> 2237</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l02238"></a><span class="lineno"> 2238</span>&#160;    TestResizeQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l02239"></a><span class="lineno"> 2239</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l02240"></a><span class="lineno"> 2240</span>&#160;</div><div class="line"><a name="l02241"></a><span class="lineno"> 2241</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l02242"></a><span class="lineno"> 2242</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02243"></a><span class="lineno"> 2243</span>&#160;    TestResizeQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l02244"></a><span class="lineno"> 2244</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l02245"></a><span class="lineno"> 2245</span>&#160;</div><div class="line"><a name="l02246"></a><span class="lineno"> 2246</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l02247"></a><span class="lineno"> 2247</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02248"></a><span class="lineno"> 2248</span>&#160;    TestResizeQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l02249"></a><span class="lineno"> 2249</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l02250"></a><span class="lineno"> 2250</span>&#160;</div><div class="line"><a name="l02251"></a><span class="lineno"> 2251</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l02252"></a><span class="lineno"> 2252</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l02253"></a><span class="lineno"> 2253</span>&#160;    TestResizeQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l02254"></a><span class="lineno"> 2254</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l02255"></a><span class="lineno"> 2255</span>&#160;}</div><div class="line"><a name="l02256"></a><span class="lineno"> 2256</span>&#160;</div><div class="line"><a name="l02257"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a5f9c6094ae666c8e14907307d0481fac"> 2257</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeStridedSlice)</div><div class="line"><a name="l02258"></a><span class="lineno"> 2258</span>&#160;{</div><div class="line"><a name="l02259"></a><span class="lineno"> 2259</span>&#160;    <span class="keyword">class </span>TestStridedSliceQuantization : <span class="keyword">public</span> TestLeakyReLuActivationQuantization</div><div class="line"><a name="l02260"></a><span class="lineno"> 2260</span>&#160;    {</div><div class="line"><a name="l02261"></a><span class="lineno"> 2261</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l02262"></a><span class="lineno"> 2262</span>&#160;        TestStridedSliceQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02263"></a><span class="lineno"> 2263</span>&#160;        : TestLeakyReLuActivationQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l02264"></a><span class="lineno"> 2264</span>&#160;</div><div class="line"><a name="l02265"></a><span class="lineno"> 2265</span>&#160;        TestStridedSliceQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l02266"></a><span class="lineno"> 2266</span>&#160;                                     <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l02267"></a><span class="lineno"> 2267</span>&#160;                                     <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02268"></a><span class="lineno"> 2268</span>&#160;        : TestLeakyReLuActivationQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l02269"></a><span class="lineno"> 2269</span>&#160;</div><div class="line"><a name="l02270"></a><span class="lineno"> 2270</span>&#160;        <span class="keyword">virtual</span> <span class="keywordtype">void</span> VisitStridedSliceLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02271"></a><span class="lineno"> 2271</span>&#160;                                            <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_strided_slice_descriptor.html">StridedSliceDescriptor</a>&amp; desc,</div><div class="line"><a name="l02272"></a><span class="lineno"> 2272</span>&#160;                                            <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)</div><div class="line"><a name="l02273"></a><span class="lineno"> 2273</span>&#160;        {</div><div class="line"><a name="l02274"></a><span class="lineno"> 2274</span>&#160;            boost::ignore_unused(desc, name);</div><div class="line"><a name="l02275"></a><span class="lineno"> 2275</span>&#160;            CheckForwardedQuantizationSettings(layer);</div><div class="line"><a name="l02276"></a><span class="lineno"> 2276</span>&#160;        }</div><div class="line"><a name="l02277"></a><span class="lineno"> 2277</span>&#160;    };</div><div class="line"><a name="l02278"></a><span class="lineno"> 2278</span>&#160;</div><div class="line"><a name="l02279"></a><span class="lineno"> 2279</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l02280"></a><span class="lineno"> 2280</span>&#160;</div><div class="line"><a name="l02281"></a><span class="lineno"> 2281</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{3U};</div><div class="line"><a name="l02282"></a><span class="lineno"> 2282</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l02283"></a><span class="lineno"> 2283</span>&#160;</div><div class="line"><a name="l02284"></a><span class="lineno"> 2284</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* activation = <a class="code" href="namespacearmnn.html#a120c131df35d78b3a56cb0f07decaf35">CreateStartOfLeakyReluNetwork</a>(network.get(), <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l02285"></a><span class="lineno"> 2285</span>&#160;</div><div class="line"><a name="l02286"></a><span class="lineno"> 2286</span>&#160;    <span class="comment">// Add the layer under test</span></div><div class="line"><a name="l02287"></a><span class="lineno"> 2287</span>&#160;    <a class="code" href="structarmnn_1_1_strided_slice_descriptor.html">StridedSliceDescriptor</a> stridedSliceDesc;</div><div class="line"><a name="l02288"></a><span class="lineno"> 2288</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* stridedSlice = network-&gt;AddStridedSliceLayer(stridedSliceDesc);</div><div class="line"><a name="l02289"></a><span class="lineno"> 2289</span>&#160;</div><div class="line"><a name="l02290"></a><span class="lineno"> 2290</span>&#160;    <a class="code" href="namespacearmnn.html#a6fff4b4b1b5d4d37c9cf53d0e31c05dd">CompleteLeakyReluNetwork</a>(network.get(), activation, stridedSlice, <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l02291"></a><span class="lineno"> 2291</span>&#160;</div><div class="line"><a name="l02292"></a><span class="lineno"> 2292</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l02293"></a><span class="lineno"> 2293</span>&#160;    TestStridedSliceQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l02294"></a><span class="lineno"> 2294</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l02295"></a><span class="lineno"> 2295</span>&#160;</div><div class="line"><a name="l02296"></a><span class="lineno"> 2296</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l02297"></a><span class="lineno"> 2297</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02298"></a><span class="lineno"> 2298</span>&#160;    TestStridedSliceQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l02299"></a><span class="lineno"> 2299</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l02300"></a><span class="lineno"> 2300</span>&#160;</div><div class="line"><a name="l02301"></a><span class="lineno"> 2301</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l02302"></a><span class="lineno"> 2302</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02303"></a><span class="lineno"> 2303</span>&#160;    TestStridedSliceQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l02304"></a><span class="lineno"> 2304</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l02305"></a><span class="lineno"> 2305</span>&#160;</div><div class="line"><a name="l02306"></a><span class="lineno"> 2306</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l02307"></a><span class="lineno"> 2307</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l02308"></a><span class="lineno"> 2308</span>&#160;    TestStridedSliceQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l02309"></a><span class="lineno"> 2309</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l02310"></a><span class="lineno"> 2310</span>&#160;}</div><div class="line"><a name="l02311"></a><span class="lineno"> 2311</span>&#160;</div><div class="line"><a name="l02312"></a><span class="lineno"><a class="line" href="namespacearmnn.html#aec7cf8e3927ee7d24f8b19d206ce3e84"> 2312</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeBatchToSpace)</div><div class="line"><a name="l02313"></a><span class="lineno"> 2313</span>&#160;{</div><div class="line"><a name="l02314"></a><span class="lineno"> 2314</span>&#160;    <span class="keyword">class </span>TestBatchToSpaceQuantization : <span class="keyword">public</span> TestLeakyReLuActivationQuantization</div><div class="line"><a name="l02315"></a><span class="lineno"> 2315</span>&#160;    {</div><div class="line"><a name="l02316"></a><span class="lineno"> 2316</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l02317"></a><span class="lineno"> 2317</span>&#160;        TestBatchToSpaceQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02318"></a><span class="lineno"> 2318</span>&#160;        : TestLeakyReLuActivationQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l02319"></a><span class="lineno"> 2319</span>&#160;</div><div class="line"><a name="l02320"></a><span class="lineno"> 2320</span>&#160;        TestBatchToSpaceQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l02321"></a><span class="lineno"> 2321</span>&#160;                                     <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l02322"></a><span class="lineno"> 2322</span>&#160;                                     <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02323"></a><span class="lineno"> 2323</span>&#160;        : TestLeakyReLuActivationQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l02324"></a><span class="lineno"> 2324</span>&#160;</div><div class="line"><a name="l02325"></a><span class="lineno"> 2325</span>&#160;        <span class="keywordtype">void</span> VisitBatchToSpaceNdLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02326"></a><span class="lineno"> 2326</span>&#160;                                      <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_batch_to_space_nd_descriptor.html">BatchToSpaceNdDescriptor</a>&amp; batchToSpaceNdDescriptor,</div><div class="line"><a name="l02327"></a><span class="lineno"> 2327</span>&#160;                                      <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02328"></a><span class="lineno"> 2328</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l02329"></a><span class="lineno"> 2329</span>&#160;            boost::ignore_unused(batchToSpaceNdDescriptor, name);</div><div class="line"><a name="l02330"></a><span class="lineno"> 2330</span>&#160;            CheckForwardedQuantizationSettings(layer);</div><div class="line"><a name="l02331"></a><span class="lineno"> 2331</span>&#160;        }</div><div class="line"><a name="l02332"></a><span class="lineno"> 2332</span>&#160;    };</div><div class="line"><a name="l02333"></a><span class="lineno"> 2333</span>&#160;</div><div class="line"><a name="l02334"></a><span class="lineno"> 2334</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l02335"></a><span class="lineno"> 2335</span>&#160;</div><div class="line"><a name="l02336"></a><span class="lineno"> 2336</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U};</div><div class="line"><a name="l02337"></a><span class="lineno"> 2337</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l02338"></a><span class="lineno"> 2338</span>&#160;</div><div class="line"><a name="l02339"></a><span class="lineno"> 2339</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* activation = <a class="code" href="namespacearmnn.html#a120c131df35d78b3a56cb0f07decaf35">CreateStartOfLeakyReluNetwork</a>(network.get(), <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l02340"></a><span class="lineno"> 2340</span>&#160;</div><div class="line"><a name="l02341"></a><span class="lineno"> 2341</span>&#160;    <span class="comment">// Add the layer under test</span></div><div class="line"><a name="l02342"></a><span class="lineno"> 2342</span>&#160;    <a class="code" href="structarmnn_1_1_batch_to_space_nd_descriptor.html">BatchToSpaceNdDescriptor</a> descriptor;</div><div class="line"><a name="l02343"></a><span class="lineno"> 2343</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* batchToSpace = network-&gt;AddBatchToSpaceNdLayer(descriptor);</div><div class="line"><a name="l02344"></a><span class="lineno"> 2344</span>&#160;</div><div class="line"><a name="l02345"></a><span class="lineno"> 2345</span>&#160;    <a class="code" href="namespacearmnn.html#a6fff4b4b1b5d4d37c9cf53d0e31c05dd">CompleteLeakyReluNetwork</a>(network.get(), activation, batchToSpace, <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>);</div><div class="line"><a name="l02346"></a><span class="lineno"> 2346</span>&#160;</div><div class="line"><a name="l02347"></a><span class="lineno"> 2347</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l02348"></a><span class="lineno"> 2348</span>&#160;    TestBatchToSpaceQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l02349"></a><span class="lineno"> 2349</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l02350"></a><span class="lineno"> 2350</span>&#160;</div><div class="line"><a name="l02351"></a><span class="lineno"> 2351</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l02352"></a><span class="lineno"> 2352</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02353"></a><span class="lineno"> 2353</span>&#160;    TestBatchToSpaceQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l02354"></a><span class="lineno"> 2354</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l02355"></a><span class="lineno"> 2355</span>&#160;</div><div class="line"><a name="l02356"></a><span class="lineno"> 2356</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l02357"></a><span class="lineno"> 2357</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02358"></a><span class="lineno"> 2358</span>&#160;    TestBatchToSpaceQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l02359"></a><span class="lineno"> 2359</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l02360"></a><span class="lineno"> 2360</span>&#160;</div><div class="line"><a name="l02361"></a><span class="lineno"> 2361</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l02362"></a><span class="lineno"> 2362</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l02363"></a><span class="lineno"> 2363</span>&#160;    TestBatchToSpaceQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l02364"></a><span class="lineno"> 2364</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l02365"></a><span class="lineno"> 2365</span>&#160;}</div><div class="line"><a name="l02366"></a><span class="lineno"> 2366</span>&#160;</div><div class="line"><a name="l02367"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a733ef16d4eaaf8cce338320fa042f526"> 2367</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizePrelu)</div><div class="line"><a name="l02368"></a><span class="lineno"> 2368</span>&#160;{</div><div class="line"><a name="l02369"></a><span class="lineno"> 2369</span>&#160;    <span class="keyword">class </span>TestPreluQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l02370"></a><span class="lineno"> 2370</span>&#160;    {</div><div class="line"><a name="l02371"></a><span class="lineno"> 2371</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l02372"></a><span class="lineno"> 2372</span>&#160;        TestPreluQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l02373"></a><span class="lineno"> 2373</span>&#160;                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; alphaShape,</div><div class="line"><a name="l02374"></a><span class="lineno"> 2374</span>&#160;                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02375"></a><span class="lineno"> 2375</span>&#160;            : TestQuantization(inputShape, outputShape)</div><div class="line"><a name="l02376"></a><span class="lineno"> 2376</span>&#160;            , m_AlphaShape(alphaShape)</div><div class="line"><a name="l02377"></a><span class="lineno"> 2377</span>&#160;        {}</div><div class="line"><a name="l02378"></a><span class="lineno"> 2378</span>&#160;</div><div class="line"><a name="l02379"></a><span class="lineno"> 2379</span>&#160;        TestPreluQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l02380"></a><span class="lineno"> 2380</span>&#160;                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l02381"></a><span class="lineno"> 2381</span>&#160;                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; alphaShape,</div><div class="line"><a name="l02382"></a><span class="lineno"> 2382</span>&#160;                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02383"></a><span class="lineno"> 2383</span>&#160;            : TestQuantization(options, inputShape, outputShape)</div><div class="line"><a name="l02384"></a><span class="lineno"> 2384</span>&#160;            , m_AlphaShape(alphaShape)</div><div class="line"><a name="l02385"></a><span class="lineno"> 2385</span>&#160;        {}</div><div class="line"><a name="l02386"></a><span class="lineno"> 2386</span>&#160;</div><div class="line"><a name="l02387"></a><span class="lineno"> 2387</span>&#160;        <span class="keywordtype">void</span> VisitInputLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02388"></a><span class="lineno"> 2388</span>&#160;                             <a class="code" href="namespacearmnn.html#ab8cf8f9fb6792e654c2d8d8382f6f01b">LayerBindingId</a> <span class="keywordtype">id</span>,</div><div class="line"><a name="l02389"></a><span class="lineno"> 2389</span>&#160;                             <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02390"></a><span class="lineno"> 2390</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l02391"></a><span class="lineno"> 2391</span>&#160;            boost::ignore_unused(<span class="keywordtype">id</span>, name);</div><div class="line"><a name="l02392"></a><span class="lineno"> 2392</span>&#160;            <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l02393"></a><span class="lineno"> 2393</span>&#160;</div><div class="line"><a name="l02394"></a><span class="lineno"> 2394</span>&#160;            <span class="keywordflow">switch</span> (<span class="keywordtype">id</span>)</div><div class="line"><a name="l02395"></a><span class="lineno"> 2395</span>&#160;            {</div><div class="line"><a name="l02396"></a><span class="lineno"> 2396</span>&#160;            <span class="keywordflow">case</span> 0: <span class="comment">// Input</span></div><div class="line"><a name="l02397"></a><span class="lineno"> 2397</span>&#160;                BOOST_TEST(m_InputShape == info.<a class="code" href="classarmnn_1_1_tensor_info.html#a8b5d0f8a24e9d9238f412260a552acf8">GetShape</a>());</div><div class="line"><a name="l02398"></a><span class="lineno"> 2398</span>&#160;                <span class="keywordflow">break</span>;</div><div class="line"><a name="l02399"></a><span class="lineno"> 2399</span>&#160;            <span class="keywordflow">case</span> 1: <span class="comment">// Alpha</span></div><div class="line"><a name="l02400"></a><span class="lineno"> 2400</span>&#160;                BOOST_TEST(m_AlphaShape == info.<a class="code" href="classarmnn_1_1_tensor_info.html#a8b5d0f8a24e9d9238f412260a552acf8">GetShape</a>());</div><div class="line"><a name="l02401"></a><span class="lineno"> 2401</span>&#160;                <span class="keywordflow">break</span>;</div><div class="line"><a name="l02402"></a><span class="lineno"> 2402</span>&#160;            <span class="keywordflow">default</span>:</div><div class="line"><a name="l02403"></a><span class="lineno"> 2403</span>&#160;                <span class="keywordflow">throw</span> <a class="code" href="classarmnn_1_1_invalid_argument_exception.html">InvalidArgumentException</a>(<span class="stringliteral">&quot;Invalid layer binding id for PReLU layer&quot;</span>);</div><div class="line"><a name="l02404"></a><span class="lineno"> 2404</span>&#160;            }</div><div class="line"><a name="l02405"></a><span class="lineno"> 2405</span>&#160;</div><div class="line"><a name="l02406"></a><span class="lineno"> 2406</span>&#160;            <span class="comment">// Based off current default [-15.0f, 15.0f]</span></div><div class="line"><a name="l02407"></a><span class="lineno"> 2407</span>&#160;            TestQuantizationParams(info,</div><div class="line"><a name="l02408"></a><span class="lineno"> 2408</span>&#160;                                   { 30.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 128 }, <span class="comment">// QASymmU8</span></div><div class="line"><a name="l02409"></a><span class="lineno"> 2409</span>&#160;                                   { 30.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>,  0},   <span class="comment">// QASymmS8</span></div><div class="line"><a name="l02410"></a><span class="lineno"> 2410</span>&#160;                                   { 15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0},    <span class="comment">// QSymmS8</span></div><div class="line"><a name="l02411"></a><span class="lineno"> 2411</span>&#160;                                   { 15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0 });  <span class="comment">// QSymmS16</span></div><div class="line"><a name="l02412"></a><span class="lineno"> 2412</span>&#160;        }</div><div class="line"><a name="l02413"></a><span class="lineno"> 2413</span>&#160;</div><div class="line"><a name="l02414"></a><span class="lineno"> 2414</span>&#160;        <span class="keywordtype">void</span> VisitOutputLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02415"></a><span class="lineno"> 2415</span>&#160;                              <a class="code" href="namespacearmnn.html#ab8cf8f9fb6792e654c2d8d8382f6f01b">LayerBindingId</a> <span class="keywordtype">id</span>,</div><div class="line"><a name="l02416"></a><span class="lineno"> 2416</span>&#160;                              <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02417"></a><span class="lineno"> 2417</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l02418"></a><span class="lineno"> 2418</span>&#160;            boost::ignore_unused(<span class="keywordtype">id</span>, name);</div><div class="line"><a name="l02419"></a><span class="lineno"> 2419</span>&#160;            <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_input_slot.html#a81fbf6103761e55061b62ba989b00f10">GetConnection</a>()-&gt;<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l02420"></a><span class="lineno"> 2420</span>&#160;            BOOST_TEST(m_OutputShape == info.<a class="code" href="classarmnn_1_1_tensor_info.html#a8b5d0f8a24e9d9238f412260a552acf8">GetShape</a>());</div><div class="line"><a name="l02421"></a><span class="lineno"> 2421</span>&#160;        }</div><div class="line"><a name="l02422"></a><span class="lineno"> 2422</span>&#160;</div><div class="line"><a name="l02423"></a><span class="lineno"> 2423</span>&#160;        <span class="keywordtype">void</span> VisitPreluLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02424"></a><span class="lineno"> 2424</span>&#160;                             <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02425"></a><span class="lineno"> 2425</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l02426"></a><span class="lineno"> 2426</span>&#160;            boost::ignore_unused(name);</div><div class="line"><a name="l02427"></a><span class="lineno"> 2427</span>&#160;            <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l02428"></a><span class="lineno"> 2428</span>&#160;            TestQuantizationParams(info,</div><div class="line"><a name="l02429"></a><span class="lineno"> 2429</span>&#160;                                   { 30.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 128 }, <span class="comment">// QASymmU8</span></div><div class="line"><a name="l02430"></a><span class="lineno"> 2430</span>&#160;                                   { 30.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>,  0},   <span class="comment">// QAsymmS8</span></div><div class="line"><a name="l02431"></a><span class="lineno"> 2431</span>&#160;                                   { 15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0},    <span class="comment">// QSymmS8</span></div><div class="line"><a name="l02432"></a><span class="lineno"> 2432</span>&#160;                                   { 15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0 });  <span class="comment">// QSymmS16</span></div><div class="line"><a name="l02433"></a><span class="lineno"> 2433</span>&#160;        }</div><div class="line"><a name="l02434"></a><span class="lineno"> 2434</span>&#160;</div><div class="line"><a name="l02435"></a><span class="lineno"> 2435</span>&#160;    <span class="keyword">private</span>:</div><div class="line"><a name="l02436"></a><span class="lineno"> 2436</span>&#160;        <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> m_AlphaShape;</div><div class="line"><a name="l02437"></a><span class="lineno"> 2437</span>&#160;    };</div><div class="line"><a name="l02438"></a><span class="lineno"> 2438</span>&#160;</div><div class="line"><a name="l02439"></a><span class="lineno"> 2439</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l02440"></a><span class="lineno"> 2440</span>&#160;</div><div class="line"><a name="l02441"></a><span class="lineno"> 2441</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> inputShape{ 4, 1, 2 };</div><div class="line"><a name="l02442"></a><span class="lineno"> 2442</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> alphaShape{ 5, 4, 3, 1 };</div><div class="line"><a name="l02443"></a><span class="lineno"> 2443</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> outputShape{ 5, 4, 3, 2 };</div><div class="line"><a name="l02444"></a><span class="lineno"> 2444</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> inputInfo(inputShape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l02445"></a><span class="lineno"> 2445</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> alphaInfo(alphaShape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l02446"></a><span class="lineno"> 2446</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> outputInfo(outputShape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l02447"></a><span class="lineno"> 2447</span>&#160;</div><div class="line"><a name="l02448"></a><span class="lineno"> 2448</span>&#160;    <span class="comment">// Add the input layers</span></div><div class="line"><a name="l02449"></a><span class="lineno"> 2449</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l02450"></a><span class="lineno"> 2450</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* alpha = network-&gt;AddInputLayer(1);</div><div class="line"><a name="l02451"></a><span class="lineno"> 2451</span>&#160;</div><div class="line"><a name="l02452"></a><span class="lineno"> 2452</span>&#160;    <span class="comment">// Add the layer under test</span></div><div class="line"><a name="l02453"></a><span class="lineno"> 2453</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* prelu = network-&gt;AddPreluLayer(<span class="stringliteral">&quot;prelu&quot;</span>);</div><div class="line"><a name="l02454"></a><span class="lineno"> 2454</span>&#160;</div><div class="line"><a name="l02455"></a><span class="lineno"> 2455</span>&#160;    <span class="comment">// Add the output layers</span></div><div class="line"><a name="l02456"></a><span class="lineno"> 2456</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;AddOutputLayer(0);</div><div class="line"><a name="l02457"></a><span class="lineno"> 2457</span>&#160;</div><div class="line"><a name="l02458"></a><span class="lineno"> 2458</span>&#160;    <span class="comment">// Establish connections</span></div><div class="line"><a name="l02459"></a><span class="lineno"> 2459</span>&#160;    input-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(prelu-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02460"></a><span class="lineno"> 2460</span>&#160;    alpha-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(prelu-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(1));</div><div class="line"><a name="l02461"></a><span class="lineno"> 2461</span>&#160;    prelu-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02462"></a><span class="lineno"> 2462</span>&#160;</div><div class="line"><a name="l02463"></a><span class="lineno"> 2463</span>&#160;    <span class="comment">// Set tensor info</span></div><div class="line"><a name="l02464"></a><span class="lineno"> 2464</span>&#160;    input-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(inputInfo);</div><div class="line"><a name="l02465"></a><span class="lineno"> 2465</span>&#160;    alpha-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(alphaInfo);</div><div class="line"><a name="l02466"></a><span class="lineno"> 2466</span>&#160;    prelu-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(outputInfo);</div><div class="line"><a name="l02467"></a><span class="lineno"> 2467</span>&#160;</div><div class="line"><a name="l02468"></a><span class="lineno"> 2468</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l02469"></a><span class="lineno"> 2469</span>&#160;    TestPreluQuantization validatorQAsymmU8(inputShape, alphaShape, outputShape);</div><div class="line"><a name="l02470"></a><span class="lineno"> 2470</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l02471"></a><span class="lineno"> 2471</span>&#160;</div><div class="line"><a name="l02472"></a><span class="lineno"> 2472</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l02473"></a><span class="lineno"> 2473</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02474"></a><span class="lineno"> 2474</span>&#160;    TestPreluQuantization validatorQAsymmS8(qAsymmS8Options, inputShape, alphaShape, outputShape);</div><div class="line"><a name="l02475"></a><span class="lineno"> 2475</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l02476"></a><span class="lineno"> 2476</span>&#160;</div><div class="line"><a name="l02477"></a><span class="lineno"> 2477</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l02478"></a><span class="lineno"> 2478</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02479"></a><span class="lineno"> 2479</span>&#160;    TestPreluQuantization validatorQSymmS8(qSymmS8Options, inputShape, alphaShape, outputShape);</div><div class="line"><a name="l02480"></a><span class="lineno"> 2480</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l02481"></a><span class="lineno"> 2481</span>&#160;</div><div class="line"><a name="l02482"></a><span class="lineno"> 2482</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l02483"></a><span class="lineno"> 2483</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l02484"></a><span class="lineno"> 2484</span>&#160;    TestPreluQuantization validatorQSymmS16(qSymmS16options, inputShape, alphaShape, outputShape);</div><div class="line"><a name="l02485"></a><span class="lineno"> 2485</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l02486"></a><span class="lineno"> 2486</span>&#160;}</div><div class="line"><a name="l02487"></a><span class="lineno"> 2487</span>&#160;</div><div class="line"><a name="l02488"></a><span class="lineno"><a class="line" href="namespacearmnn.html#afa7a0a639e2772ff2ced67d77be810c0"> 2488</a></span>&#160;<span class="keywordtype">void</span> <a class="code" href="namespacearmnn.html#afa7a0a639e2772ff2ced67d77be810c0">TestQuantizeTransposeConvolution2d</a>(<span class="keywordtype">bool</span> useBiases)</div><div class="line"><a name="l02489"></a><span class="lineno"> 2489</span>&#160;{</div><div class="line"><a name="l02490"></a><span class="lineno"> 2490</span>&#160;    <span class="keyword">class </span>TestTransposeConvolution2dQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l02491"></a><span class="lineno"> 2491</span>&#160;    {</div><div class="line"><a name="l02492"></a><span class="lineno"> 2492</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l02493"></a><span class="lineno"> 2493</span>&#160;        TestTransposeConvolution2dQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape) :</div><div class="line"><a name="l02494"></a><span class="lineno"> 2494</span>&#160;            TestQuantization(inputShape, outputShape)</div><div class="line"><a name="l02495"></a><span class="lineno"> 2495</span>&#160;        {}</div><div class="line"><a name="l02496"></a><span class="lineno"> 2496</span>&#160;</div><div class="line"><a name="l02497"></a><span class="lineno"> 2497</span>&#160;        TestTransposeConvolution2dQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l02498"></a><span class="lineno"> 2498</span>&#160;                                               <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l02499"></a><span class="lineno"> 2499</span>&#160;                                               <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape) :</div><div class="line"><a name="l02500"></a><span class="lineno"> 2500</span>&#160;            TestQuantization(options, inputShape, outputShape)</div><div class="line"><a name="l02501"></a><span class="lineno"> 2501</span>&#160;        {}</div><div class="line"><a name="l02502"></a><span class="lineno"> 2502</span>&#160;</div><div class="line"><a name="l02503"></a><span class="lineno"> 2503</span>&#160;        <span class="keywordtype">void</span> VisitTransposeConvolution2dLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a> *layer,</div><div class="line"><a name="l02504"></a><span class="lineno"> 2504</span>&#160;                                              <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_transpose_convolution2d_descriptor.html">TransposeConvolution2dDescriptor</a>&amp; descriptor,</div><div class="line"><a name="l02505"></a><span class="lineno"> 2505</span>&#160;                                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a>&amp; weights,</div><div class="line"><a name="l02506"></a><span class="lineno"> 2506</span>&#160;                                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_optional.html">Optional&lt;ConstTensor&gt;</a>&amp; biases,</div><div class="line"><a name="l02507"></a><span class="lineno"> 2507</span>&#160;                                              <span class="keyword">const</span> <span class="keywordtype">char</span> *name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02508"></a><span class="lineno"> 2508</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l02509"></a><span class="lineno"> 2509</span>&#160;            boost::ignore_unused(descriptor, name);</div><div class="line"><a name="l02510"></a><span class="lineno"> 2510</span>&#160;            TestQuantizationOnLayersWithBiases(layer, weights, biases);</div><div class="line"><a name="l02511"></a><span class="lineno"> 2511</span>&#160;        }</div><div class="line"><a name="l02512"></a><span class="lineno"> 2512</span>&#160;    };</div><div class="line"><a name="l02513"></a><span class="lineno"> 2513</span>&#160;</div><div class="line"><a name="l02514"></a><span class="lineno"> 2514</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l02515"></a><span class="lineno"> 2515</span>&#160;</div><div class="line"><a name="l02516"></a><span class="lineno"> 2516</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{ 3 };</div><div class="line"><a name="l02517"></a><span class="lineno"> 2517</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l02518"></a><span class="lineno"> 2518</span>&#160;</div><div class="line"><a name="l02519"></a><span class="lineno"> 2519</span>&#160;    std::initializer_list&lt;float&gt; floatData{ -1.0f, 1.5f, 2.0f };</div><div class="line"><a name="l02520"></a><span class="lineno"> 2520</span>&#160;    std::vector&lt;float&gt; weightsData(floatData);</div><div class="line"><a name="l02521"></a><span class="lineno"> 2521</span>&#160;    <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a> weights(info, weightsData);</div><div class="line"><a name="l02522"></a><span class="lineno"> 2522</span>&#160;</div><div class="line"><a name="l02523"></a><span class="lineno"> 2523</span>&#160;    <a class="code" href="structarmnn_1_1_transpose_convolution2d_descriptor.html">TransposeConvolution2dDescriptor</a> descriptor;</div><div class="line"><a name="l02524"></a><span class="lineno"> 2524</span>&#160;    descriptor.<a class="code" href="structarmnn_1_1_transpose_convolution2d_descriptor.html#aea202e14d8874cefd9a0f778022b7e25">m_BiasEnabled</a> = useBiases;</div><div class="line"><a name="l02525"></a><span class="lineno"> 2525</span>&#160;</div><div class="line"><a name="l02526"></a><span class="lineno"> 2526</span>&#160;    <span class="comment">// construct network</span></div><div class="line"><a name="l02527"></a><span class="lineno"> 2527</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l02528"></a><span class="lineno"> 2528</span>&#160;    <a class="code" href="classarmnn_1_1_optional.html">Optional&lt;ConstTensor&gt;</a> optionalBiases;</div><div class="line"><a name="l02529"></a><span class="lineno"> 2529</span>&#160;    std::vector&lt;float&gt; biasesData(floatData);</div><div class="line"><a name="l02530"></a><span class="lineno"> 2530</span>&#160;    <span class="keywordflow">if</span> (useBiases)</div><div class="line"><a name="l02531"></a><span class="lineno"> 2531</span>&#160;    {</div><div class="line"><a name="l02532"></a><span class="lineno"> 2532</span>&#160;        <a class="code" href="classarmnn_1_1_const_tensor.html">ConstTensor</a> biases(info, biasesData);</div><div class="line"><a name="l02533"></a><span class="lineno"> 2533</span>&#160;        optionalBiases = <a class="code" href="classarmnn_1_1_optional.html">Optional&lt;ConstTensor&gt;</a>(biases);</div><div class="line"><a name="l02534"></a><span class="lineno"> 2534</span>&#160;    }</div><div class="line"><a name="l02535"></a><span class="lineno"> 2535</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* transposeConv2d = network-&gt;AddTransposeConvolution2dLayer(descriptor, weights, optionalBiases);</div><div class="line"><a name="l02536"></a><span class="lineno"> 2536</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;AddOutputLayer(1);</div><div class="line"><a name="l02537"></a><span class="lineno"> 2537</span>&#160;</div><div class="line"><a name="l02538"></a><span class="lineno"> 2538</span>&#160;    input-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(transposeConv2d-&gt;GetInputSlot(0));</div><div class="line"><a name="l02539"></a><span class="lineno"> 2539</span>&#160;    transposeConv2d-&gt;GetOutputSlot(0).Connect(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02540"></a><span class="lineno"> 2540</span>&#160;</div><div class="line"><a name="l02541"></a><span class="lineno"> 2541</span>&#160;    input-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l02542"></a><span class="lineno"> 2542</span>&#160;    transposeConv2d-&gt;GetOutputSlot(0).SetTensorInfo(info);</div><div class="line"><a name="l02543"></a><span class="lineno"> 2543</span>&#160;</div><div class="line"><a name="l02544"></a><span class="lineno"> 2544</span>&#160;    <span class="comment">// test QAsymmU8 quantization</span></div><div class="line"><a name="l02545"></a><span class="lineno"> 2545</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l02546"></a><span class="lineno"> 2546</span>&#160;    TestTransposeConvolution2dQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l02547"></a><span class="lineno"> 2547</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l02548"></a><span class="lineno"> 2548</span>&#160;</div><div class="line"><a name="l02549"></a><span class="lineno"> 2549</span>&#160;    <span class="comment">//test QAsymmS8 quantization</span></div><div class="line"><a name="l02550"></a><span class="lineno"> 2550</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l02551"></a><span class="lineno"> 2551</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02552"></a><span class="lineno"> 2552</span>&#160;    TestTransposeConvolution2dQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l02553"></a><span class="lineno"> 2553</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l02554"></a><span class="lineno"> 2554</span>&#160;</div><div class="line"><a name="l02555"></a><span class="lineno"> 2555</span>&#160;    <span class="comment">// test QSymmS8 quantization</span></div><div class="line"><a name="l02556"></a><span class="lineno"> 2556</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l02557"></a><span class="lineno"> 2557</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02558"></a><span class="lineno"> 2558</span>&#160;    TestTransposeConvolution2dQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l02559"></a><span class="lineno"> 2559</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l02560"></a><span class="lineno"> 2560</span>&#160;</div><div class="line"><a name="l02561"></a><span class="lineno"> 2561</span>&#160;    <span class="comment">// test QSymmS16 quantization</span></div><div class="line"><a name="l02562"></a><span class="lineno"> 2562</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l02563"></a><span class="lineno"> 2563</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l02564"></a><span class="lineno"> 2564</span>&#160;    TestTransposeConvolution2dQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l02565"></a><span class="lineno"> 2565</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l02566"></a><span class="lineno"> 2566</span>&#160;}</div><div class="line"><a name="l02567"></a><span class="lineno"> 2567</span>&#160;</div><div class="line"><a name="l02568"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a5e66fe270ca921faeecd26735192d08b"> 2568</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeTransposeConvolution2d)</div><div class="line"><a name="l02569"></a><span class="lineno"> 2569</span>&#160;{</div><div class="line"><a name="l02570"></a><span class="lineno"> 2570</span>&#160;    <a class="code" href="namespacearmnn.html#afa7a0a639e2772ff2ced67d77be810c0">TestQuantizeTransposeConvolution2d</a>(<span class="keyword">false</span>);</div><div class="line"><a name="l02571"></a><span class="lineno"> 2571</span>&#160;}</div><div class="line"><a name="l02572"></a><span class="lineno"> 2572</span>&#160;</div><div class="line"><a name="l02573"></a><span class="lineno"><a class="line" href="namespacearmnn.html#aec82007c45313f59d24b304e35b3db6c"> 2573</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeTransposeConvolution2dWithBiases)</div><div class="line"><a name="l02574"></a><span class="lineno"> 2574</span>&#160;{</div><div class="line"><a name="l02575"></a><span class="lineno"> 2575</span>&#160;    <a class="code" href="namespacearmnn.html#afa7a0a639e2772ff2ced67d77be810c0">TestQuantizeTransposeConvolution2d</a>(<span class="keyword">true</span>);</div><div class="line"><a name="l02576"></a><span class="lineno"> 2576</span>&#160;}</div><div class="line"><a name="l02577"></a><span class="lineno"> 2577</span>&#160;</div><div class="line"><a name="l02578"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a77cba79eef903eb3d758b4edbcc626ef"> 2578</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeStack)</div><div class="line"><a name="l02579"></a><span class="lineno"> 2579</span>&#160;{</div><div class="line"><a name="l02580"></a><span class="lineno"> 2580</span>&#160;    <span class="keyword">class </span>TestStackQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l02581"></a><span class="lineno"> 2581</span>&#160;    {</div><div class="line"><a name="l02582"></a><span class="lineno"> 2582</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l02583"></a><span class="lineno"> 2583</span>&#160;        TestStackQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l02584"></a><span class="lineno"> 2584</span>&#160;                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02585"></a><span class="lineno"> 2585</span>&#160;            : TestQuantization(inputShape, outputShape) {}</div><div class="line"><a name="l02586"></a><span class="lineno"> 2586</span>&#160;</div><div class="line"><a name="l02587"></a><span class="lineno"> 2587</span>&#160;        TestStackQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l02588"></a><span class="lineno"> 2588</span>&#160;                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l02589"></a><span class="lineno"> 2589</span>&#160;                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02590"></a><span class="lineno"> 2590</span>&#160;            : TestQuantization(options, inputShape, outputShape) {}</div><div class="line"><a name="l02591"></a><span class="lineno"> 2591</span>&#160;</div><div class="line"><a name="l02592"></a><span class="lineno"> 2592</span>&#160;        <span class="keywordtype">void</span> VisitInputLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02593"></a><span class="lineno"> 2593</span>&#160;                             <a class="code" href="namespacearmnn.html#ab8cf8f9fb6792e654c2d8d8382f6f01b">LayerBindingId</a> <span class="keywordtype">id</span>,</div><div class="line"><a name="l02594"></a><span class="lineno"> 2594</span>&#160;                             <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02595"></a><span class="lineno"> 2595</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l02596"></a><span class="lineno"> 2596</span>&#160;            boost::ignore_unused(layer, <span class="keywordtype">id</span>, name);</div><div class="line"><a name="l02597"></a><span class="lineno"> 2597</span>&#160;        }</div><div class="line"><a name="l02598"></a><span class="lineno"> 2598</span>&#160;        <span class="keywordtype">void</span> VisitOutputLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02599"></a><span class="lineno"> 2599</span>&#160;                              <a class="code" href="namespacearmnn.html#ab8cf8f9fb6792e654c2d8d8382f6f01b">LayerBindingId</a> <span class="keywordtype">id</span>,</div><div class="line"><a name="l02600"></a><span class="lineno"> 2600</span>&#160;                              <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02601"></a><span class="lineno"> 2601</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l02602"></a><span class="lineno"> 2602</span>&#160;            boost::ignore_unused(layer, <span class="keywordtype">id</span>, name);</div><div class="line"><a name="l02603"></a><span class="lineno"> 2603</span>&#160;        }</div><div class="line"><a name="l02604"></a><span class="lineno"> 2604</span>&#160;</div><div class="line"><a name="l02605"></a><span class="lineno"> 2605</span>&#160;        <span class="keywordtype">void</span> VisitStackLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02606"></a><span class="lineno"> 2606</span>&#160;                             <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_stack_descriptor.html">StackDescriptor</a>&amp; descriptor,</div><div class="line"><a name="l02607"></a><span class="lineno"> 2607</span>&#160;                             <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02608"></a><span class="lineno"> 2608</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l02609"></a><span class="lineno"> 2609</span>&#160;            boost::ignore_unused(descriptor, name);</div><div class="line"><a name="l02610"></a><span class="lineno"> 2610</span>&#160;            <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> outputInfo = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l02611"></a><span class="lineno"> 2611</span>&#160;</div><div class="line"><a name="l02612"></a><span class="lineno"> 2612</span>&#160;            TestQuantizationParams(outputInfo,</div><div class="line"><a name="l02613"></a><span class="lineno"> 2613</span>&#160;                { 30.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 128 },</div><div class="line"><a name="l02614"></a><span class="lineno"> 2614</span>&#160;                { 30.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>, 0},</div><div class="line"><a name="l02615"></a><span class="lineno"> 2615</span>&#160;                { 15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0},</div><div class="line"><a name="l02616"></a><span class="lineno"> 2616</span>&#160;                { 15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0 });</div><div class="line"><a name="l02617"></a><span class="lineno"> 2617</span>&#160;        }</div><div class="line"><a name="l02618"></a><span class="lineno"> 2618</span>&#160;    };</div><div class="line"><a name="l02619"></a><span class="lineno"> 2619</span>&#160;</div><div class="line"><a name="l02620"></a><span class="lineno"> 2620</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l02621"></a><span class="lineno"> 2621</span>&#160;</div><div class="line"><a name="l02622"></a><span class="lineno"> 2622</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input0 = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l02623"></a><span class="lineno"> 2623</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input1 = network-&gt;AddInputLayer(1);</div><div class="line"><a name="l02624"></a><span class="lineno"> 2624</span>&#160;</div><div class="line"><a name="l02625"></a><span class="lineno"> 2625</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> inputShape{ 3, 4, 5 };</div><div class="line"><a name="l02626"></a><span class="lineno"> 2626</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> outputShape{ 3, 4, 2, 5 };</div><div class="line"><a name="l02627"></a><span class="lineno"> 2627</span>&#160;</div><div class="line"><a name="l02628"></a><span class="lineno"> 2628</span>&#160;    <a class="code" href="structarmnn_1_1_stack_descriptor.html">StackDescriptor</a> descriptor(2, 2, inputShape);</div><div class="line"><a name="l02629"></a><span class="lineno"> 2629</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* stackLayer = network-&gt;AddStackLayer(descriptor);</div><div class="line"><a name="l02630"></a><span class="lineno"> 2630</span>&#160;</div><div class="line"><a name="l02631"></a><span class="lineno"> 2631</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;AddOutputLayer(0);</div><div class="line"><a name="l02632"></a><span class="lineno"> 2632</span>&#160;</div><div class="line"><a name="l02633"></a><span class="lineno"> 2633</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(stackLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02634"></a><span class="lineno"> 2634</span>&#160;    input1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(stackLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(1));</div><div class="line"><a name="l02635"></a><span class="lineno"> 2635</span>&#160;    stackLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02636"></a><span class="lineno"> 2636</span>&#160;</div><div class="line"><a name="l02637"></a><span class="lineno"> 2637</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l02638"></a><span class="lineno"> 2638</span>&#160;    TestStackQuantization validatorQAsymmU8(inputShape, outputShape);</div><div class="line"><a name="l02639"></a><span class="lineno"> 2639</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l02640"></a><span class="lineno"> 2640</span>&#160;</div><div class="line"><a name="l02641"></a><span class="lineno"> 2641</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l02642"></a><span class="lineno"> 2642</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02643"></a><span class="lineno"> 2643</span>&#160;    TestStackQuantization validatorQAsymmS8(qAsymmS8Options, inputShape, inputShape);</div><div class="line"><a name="l02644"></a><span class="lineno"> 2644</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l02645"></a><span class="lineno"> 2645</span>&#160;</div><div class="line"><a name="l02646"></a><span class="lineno"> 2646</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l02647"></a><span class="lineno"> 2647</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02648"></a><span class="lineno"> 2648</span>&#160;    TestStackQuantization validatorQSymmS8(qSymmS8Options, inputShape, inputShape);</div><div class="line"><a name="l02649"></a><span class="lineno"> 2649</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l02650"></a><span class="lineno"> 2650</span>&#160;</div><div class="line"><a name="l02651"></a><span class="lineno"> 2651</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l02652"></a><span class="lineno"> 2652</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l02653"></a><span class="lineno"> 2653</span>&#160;    TestStackQuantization validatorQSymmS16(qSymmS16options, inputShape, outputShape);</div><div class="line"><a name="l02654"></a><span class="lineno"> 2654</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l02655"></a><span class="lineno"> 2655</span>&#160;}</div><div class="line"><a name="l02656"></a><span class="lineno"> 2656</span>&#160;</div><div class="line"><a name="l02657"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a46f313720b601ca97a9c2a5158814bff"> 2657</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeSlice)</div><div class="line"><a name="l02658"></a><span class="lineno"> 2658</span>&#160;{</div><div class="line"><a name="l02659"></a><span class="lineno"> 2659</span>&#160;    <span class="keyword">class </span>TestSliceQuantization : <span class="keyword">public</span> TestQuantization</div><div class="line"><a name="l02660"></a><span class="lineno"> 2660</span>&#160;    {</div><div class="line"><a name="l02661"></a><span class="lineno"> 2661</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l02662"></a><span class="lineno"> 2662</span>&#160;        TestSliceQuantization(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape, <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02663"></a><span class="lineno"> 2663</span>&#160;            : TestQuantization(inputShape, outputShape)</div><div class="line"><a name="l02664"></a><span class="lineno"> 2664</span>&#160;        {}</div><div class="line"><a name="l02665"></a><span class="lineno"> 2665</span>&#160;</div><div class="line"><a name="l02666"></a><span class="lineno"> 2666</span>&#160;        TestSliceQuantization(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l02667"></a><span class="lineno"> 2667</span>&#160;                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l02668"></a><span class="lineno"> 2668</span>&#160;                              <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02669"></a><span class="lineno"> 2669</span>&#160;            : TestQuantization(options, inputShape, outputShape)</div><div class="line"><a name="l02670"></a><span class="lineno"> 2670</span>&#160;        {}</div><div class="line"><a name="l02671"></a><span class="lineno"> 2671</span>&#160;</div><div class="line"><a name="l02672"></a><span class="lineno"> 2672</span>&#160;        <span class="keyword">virtual</span> <span class="keywordtype">void</span> VisitSliceLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02673"></a><span class="lineno"> 2673</span>&#160;                                     <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_slice_descriptor.html">SliceDescriptor</a>&amp; desc,</div><div class="line"><a name="l02674"></a><span class="lineno"> 2674</span>&#160;                                     <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)</div><div class="line"><a name="l02675"></a><span class="lineno"> 2675</span>&#160;        {</div><div class="line"><a name="l02676"></a><span class="lineno"> 2676</span>&#160;            boost::ignore_unused(desc, name);</div><div class="line"><a name="l02677"></a><span class="lineno"> 2677</span>&#160;            <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l02678"></a><span class="lineno"> 2678</span>&#160;</div><div class="line"><a name="l02679"></a><span class="lineno"> 2679</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qAsymmU8Params{ 30.0f / <a class="code" href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">g_AsymmU8QuantizationBase</a>, 128 };</div><div class="line"><a name="l02680"></a><span class="lineno"> 2680</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qAsymmS8Params{ 30.0f / <a class="code" href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">g_AsymmS8QuantizationBase</a>, 0 };</div><div class="line"><a name="l02681"></a><span class="lineno"> 2681</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qSymmS8Params { 15.0f / <a class="code" href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">g_SymmS8QuantizationBase</a>,  0 };</div><div class="line"><a name="l02682"></a><span class="lineno"> 2682</span>&#160;            <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">OffsetScalePair</a> qSymmS16Params{ 15.0f / <a class="code" href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">g_SymmS16QuantizationBase</a>, 0 };</div><div class="line"><a name="l02683"></a><span class="lineno"> 2683</span>&#160;</div><div class="line"><a name="l02684"></a><span class="lineno"> 2684</span>&#160;            TestQuantizationParams(info, qAsymmU8Params, qAsymmS8Params, qSymmS8Params, qSymmS16Params);</div><div class="line"><a name="l02685"></a><span class="lineno"> 2685</span>&#160;        }</div><div class="line"><a name="l02686"></a><span class="lineno"> 2686</span>&#160;    };</div><div class="line"><a name="l02687"></a><span class="lineno"> 2687</span>&#160;</div><div class="line"><a name="l02688"></a><span class="lineno"> 2688</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{ 3 };</div><div class="line"><a name="l02689"></a><span class="lineno"> 2689</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l02690"></a><span class="lineno"> 2690</span>&#160;</div><div class="line"><a name="l02691"></a><span class="lineno"> 2691</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l02692"></a><span class="lineno"> 2692</span>&#160;</div><div class="line"><a name="l02693"></a><span class="lineno"> 2693</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* inputLayer  = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l02694"></a><span class="lineno"> 2694</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* sliceLayer  = network-&gt;AddSliceLayer(<a class="code" href="structarmnn_1_1_slice_descriptor.html">SliceDescriptor</a>());</div><div class="line"><a name="l02695"></a><span class="lineno"> 2695</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* outputLayer = network-&gt;AddOutputLayer(0);</div><div class="line"><a name="l02696"></a><span class="lineno"> 2696</span>&#160;</div><div class="line"><a name="l02697"></a><span class="lineno"> 2697</span>&#160;    inputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(sliceLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02698"></a><span class="lineno"> 2698</span>&#160;    sliceLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(outputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02699"></a><span class="lineno"> 2699</span>&#160;</div><div class="line"><a name="l02700"></a><span class="lineno"> 2700</span>&#160;    inputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l02701"></a><span class="lineno"> 2701</span>&#160;    sliceLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l02702"></a><span class="lineno"> 2702</span>&#160;</div><div class="line"><a name="l02703"></a><span class="lineno"> 2703</span>&#160;    <span class="comment">// test QAsymmU8 quantization</span></div><div class="line"><a name="l02704"></a><span class="lineno"> 2704</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get())-&gt;ExportNetwork();</div><div class="line"><a name="l02705"></a><span class="lineno"> 2705</span>&#160;    TestSliceQuantization validatorQAsymmU8(shape, shape);</div><div class="line"><a name="l02706"></a><span class="lineno"> 2706</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l02707"></a><span class="lineno"> 2707</span>&#160;</div><div class="line"><a name="l02708"></a><span class="lineno"> 2708</span>&#160;    <span class="comment">// test QASymmS8 quantization</span></div><div class="line"><a name="l02709"></a><span class="lineno"> 2709</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qAsymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">DataType::QAsymmS8</a>);</div><div class="line"><a name="l02710"></a><span class="lineno"> 2710</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qAsymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02711"></a><span class="lineno"> 2711</span>&#160;    TestSliceQuantization validatorQAsymmS8(qAsymmS8Options, shape, shape);</div><div class="line"><a name="l02712"></a><span class="lineno"> 2712</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmS8.get(), validatorQAsymmS8);</div><div class="line"><a name="l02713"></a><span class="lineno"> 2713</span>&#160;</div><div class="line"><a name="l02714"></a><span class="lineno"> 2714</span>&#160;    <span class="comment">// test QSymmS8 quantization</span></div><div class="line"><a name="l02715"></a><span class="lineno"> 2715</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS8Options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l02716"></a><span class="lineno"> 2716</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS8Options)-&gt;ExportNetwork();</div><div class="line"><a name="l02717"></a><span class="lineno"> 2717</span>&#160;    TestSliceQuantization validatorQSymmS8(qSymmS8Options, shape, shape);</div><div class="line"><a name="l02718"></a><span class="lineno"> 2718</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS8.get(), validatorQSymmS8);</div><div class="line"><a name="l02719"></a><span class="lineno"> 2719</span>&#160;</div><div class="line"><a name="l02720"></a><span class="lineno"> 2720</span>&#160;    <span class="comment">// test QSymmS16 quantization</span></div><div class="line"><a name="l02721"></a><span class="lineno"> 2721</span>&#160;    <span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> qSymmS16options(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l02722"></a><span class="lineno"> 2722</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQSymmS16 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), qSymmS16options)-&gt;ExportNetwork();</div><div class="line"><a name="l02723"></a><span class="lineno"> 2723</span>&#160;    TestSliceQuantization validatorQSymmS16(qSymmS16options, shape, shape);</div><div class="line"><a name="l02724"></a><span class="lineno"> 2724</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQSymmS16.get(), validatorQSymmS16);</div><div class="line"><a name="l02725"></a><span class="lineno"> 2725</span>&#160;}</div><div class="line"><a name="l02726"></a><span class="lineno"> 2726</span>&#160;</div><div class="line"><a name="l02727"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a52cbff9d344ba4a1fe01d4da2c1f7ba2"> 2727</a></span>&#160;std::vector&lt;uint8_t&gt; <a class="code" href="namespacearmnn.html#a52cbff9d344ba4a1fe01d4da2c1f7ba2">SetupQuantize</a>(<span class="keywordtype">float</span> value)</div><div class="line"><a name="l02728"></a><span class="lineno"> 2728</span>&#160;{</div><div class="line"><a name="l02729"></a><span class="lineno"> 2729</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">armnn::TensorInfo</a> inputInfo({ 1, 2, 2 }, <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">armnn::DataType::Float32</a>);</div><div class="line"><a name="l02730"></a><span class="lineno"> 2730</span>&#160;    inputInfo.<a class="code" href="classarmnn_1_1_tensor_info.html#a685739c4eb65a580e075282cfe6787d6">SetQuantizationScale</a>(1.0f);</div><div class="line"><a name="l02731"></a><span class="lineno"> 2731</span>&#160;    inputInfo.SetQuantizationOffset(1);</div><div class="line"><a name="l02732"></a><span class="lineno"> 2732</span>&#160;    std::vector&lt;float&gt; input({ value, 0.0f, 0.0f, 1.0f });</div><div class="line"><a name="l02733"></a><span class="lineno"> 2733</span>&#160;    <span class="keyword">const</span> std::vector&lt;float&gt; &amp;inputRef = input;</div><div class="line"><a name="l02734"></a><span class="lineno"> 2734</span>&#160;</div><div class="line"><a name="l02735"></a><span class="lineno"> 2735</span>&#160;    <span class="keyword">auto</span> output = armnnUtils::QuantizedVector&lt;uint8_t&gt;(inputRef,</div><div class="line"><a name="l02736"></a><span class="lineno"> 2736</span>&#160;                                                       inputInfo.GetQuantizationScale(),</div><div class="line"><a name="l02737"></a><span class="lineno"> 2737</span>&#160;                                                       inputInfo.GetQuantizationOffset());</div><div class="line"><a name="l02738"></a><span class="lineno"> 2738</span>&#160;</div><div class="line"><a name="l02739"></a><span class="lineno"> 2739</span>&#160;    <span class="keywordflow">return</span> output;</div><div class="line"><a name="l02740"></a><span class="lineno"> 2740</span>&#160;}</div><div class="line"><a name="l02741"></a><span class="lineno"> 2741</span>&#160;</div><div class="line"><a name="l02742"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a728153b62fa66e6ed1243e09144bfe8c"> 2742</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeInf)</div><div class="line"><a name="l02743"></a><span class="lineno"> 2743</span>&#160;{</div><div class="line"><a name="l02744"></a><span class="lineno"> 2744</span>&#160;    BOOST_CHECK_EQUAL(<a class="code" href="namespacearmnn.html#a52cbff9d344ba4a1fe01d4da2c1f7ba2">SetupQuantize</a>(std::numeric_limits&lt;float&gt;::infinity())[0], 255);</div><div class="line"><a name="l02745"></a><span class="lineno"> 2745</span>&#160;}</div><div class="line"><a name="l02746"></a><span class="lineno"> 2746</span>&#160;</div><div class="line"><a name="l02747"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a898305dc4cdb78a5fbed481250f6cd35"> 2747</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(QuantizeNegativeInf)</div><div class="line"><a name="l02748"></a><span class="lineno"> 2748</span>&#160;{</div><div class="line"><a name="l02749"></a><span class="lineno"> 2749</span>&#160;    BOOST_CHECK_EQUAL(<a class="code" href="namespacearmnn.html#a52cbff9d344ba4a1fe01d4da2c1f7ba2">SetupQuantize</a>(-1 * std::numeric_limits&lt;float&gt;::infinity())[0], 0);</div><div class="line"><a name="l02750"></a><span class="lineno"> 2750</span>&#160;}</div><div class="line"><a name="l02751"></a><span class="lineno"> 2751</span>&#160;</div><div class="line"><a name="l02752"></a><span class="lineno"> 2752</span>&#160;<span class="keyword">class </span>TestPreserveType : <span class="keyword">public</span> TestAdditionQuantization</div><div class="line"><a name="l02753"></a><span class="lineno"> 2753</span>&#160;{</div><div class="line"><a name="l02754"></a><span class="lineno"> 2754</span>&#160;<span class="keyword">public</span>:</div><div class="line"><a name="l02755"></a><span class="lineno"> 2755</span>&#160;    TestPreserveType(<span class="keyword">const</span> <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>&amp; <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>,</div><div class="line"><a name="l02756"></a><span class="lineno"> 2756</span>&#160;                     <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6">DataType</a>&amp; dataType,</div><div class="line"><a name="l02757"></a><span class="lineno"> 2757</span>&#160;                     <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; inputShape,</div><div class="line"><a name="l02758"></a><span class="lineno"> 2758</span>&#160;                     <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>&amp; outputShape)</div><div class="line"><a name="l02759"></a><span class="lineno"> 2759</span>&#160;    : TestAdditionQuantization(options, inputShape, outputShape)</div><div class="line"><a name="l02760"></a><span class="lineno"> 2760</span>&#160;    , m_DataType(dataType)</div><div class="line"><a name="l02761"></a><span class="lineno"> 2761</span>&#160;    , m_VisitedQuantizeLayer(<span class="keyword">false</span>)</div><div class="line"><a name="l02762"></a><span class="lineno"> 2762</span>&#160;    , m_VisitedDequantizeLayer(<span class="keyword">false</span>) {}</div><div class="line"><a name="l02763"></a><span class="lineno"> 2763</span>&#160;</div><div class="line"><a name="l02764"></a><span class="lineno"> 2764</span>&#160;    <span class="keywordtype">void</span> VisitInputLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02765"></a><span class="lineno"> 2765</span>&#160;                         <a class="code" href="namespacearmnn.html#ab8cf8f9fb6792e654c2d8d8382f6f01b">LayerBindingId</a> <span class="keywordtype">id</span>,</div><div class="line"><a name="l02766"></a><span class="lineno"> 2766</span>&#160;                         <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02767"></a><span class="lineno"> 2767</span>&#160;<span class="keyword">    </span>{</div><div class="line"><a name="l02768"></a><span class="lineno"> 2768</span>&#160;        boost::ignore_unused(<span class="keywordtype">id</span>, name);</div><div class="line"><a name="l02769"></a><span class="lineno"> 2769</span>&#160;        <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l02770"></a><span class="lineno"> 2770</span>&#160;        BOOST_TEST(<a class="code" href="namespacearmnn.html#a81b5ff8545adad19a1c9d4ca076d552c">GetDataTypeName</a>(info.<a class="code" href="classarmnn_1_1_tensor_info.html#aea909c7327109228ef618d459015def3">GetDataType</a>()) == <a class="code" href="namespacearmnn.html#a81b5ff8545adad19a1c9d4ca076d552c">GetDataTypeName</a>(m_DataType));</div><div class="line"><a name="l02771"></a><span class="lineno"> 2771</span>&#160;        BOOST_TEST(m_InputShape == info.<a class="code" href="classarmnn_1_1_tensor_info.html#a8b5d0f8a24e9d9238f412260a552acf8">GetShape</a>());</div><div class="line"><a name="l02772"></a><span class="lineno"> 2772</span>&#160;    }</div><div class="line"><a name="l02773"></a><span class="lineno"> 2773</span>&#160;</div><div class="line"><a name="l02774"></a><span class="lineno"> 2774</span>&#160;    <span class="keywordtype">void</span> VisitOutputLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02775"></a><span class="lineno"> 2775</span>&#160;                          <a class="code" href="namespacearmnn.html#ab8cf8f9fb6792e654c2d8d8382f6f01b">LayerBindingId</a> <span class="keywordtype">id</span>,</div><div class="line"><a name="l02776"></a><span class="lineno"> 2776</span>&#160;                          <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02777"></a><span class="lineno"> 2777</span>&#160;<span class="keyword">    </span>{</div><div class="line"><a name="l02778"></a><span class="lineno"> 2778</span>&#160;        boost::ignore_unused(<span class="keywordtype">id</span>, name);</div><div class="line"><a name="l02779"></a><span class="lineno"> 2779</span>&#160;        <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>&amp; <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a> = layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_input_slot.html#a81fbf6103761e55061b62ba989b00f10">GetConnection</a>()-&gt;<a class="code" href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">GetTensorInfo</a>();</div><div class="line"><a name="l02780"></a><span class="lineno"> 2780</span>&#160;        BOOST_TEST(<a class="code" href="namespacearmnn.html#a81b5ff8545adad19a1c9d4ca076d552c">GetDataTypeName</a>(info.<a class="code" href="classarmnn_1_1_tensor_info.html#aea909c7327109228ef618d459015def3">GetDataType</a>()) == <a class="code" href="namespacearmnn.html#a81b5ff8545adad19a1c9d4ca076d552c">GetDataTypeName</a>(m_DataType));</div><div class="line"><a name="l02781"></a><span class="lineno"> 2781</span>&#160;        BOOST_TEST(m_OutputShape == info.<a class="code" href="classarmnn_1_1_tensor_info.html#a8b5d0f8a24e9d9238f412260a552acf8">GetShape</a>());</div><div class="line"><a name="l02782"></a><span class="lineno"> 2782</span>&#160;    }</div><div class="line"><a name="l02783"></a><span class="lineno"> 2783</span>&#160;</div><div class="line"><a name="l02784"></a><span class="lineno"> 2784</span>&#160;    <span class="keywordtype">void</span> VisitQuantizeLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02785"></a><span class="lineno"> 2785</span>&#160;                            <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02786"></a><span class="lineno"> 2786</span>&#160;<span class="keyword">    </span>{</div><div class="line"><a name="l02787"></a><span class="lineno"> 2787</span>&#160;        boost::ignore_unused(layer, name);</div><div class="line"><a name="l02788"></a><span class="lineno"> 2788</span>&#160;        m_VisitedQuantizeLayer = <span class="keyword">true</span>;</div><div class="line"><a name="l02789"></a><span class="lineno"> 2789</span>&#160;    }</div><div class="line"><a name="l02790"></a><span class="lineno"> 2790</span>&#160;</div><div class="line"><a name="l02791"></a><span class="lineno"> 2791</span>&#160;    <span class="keywordtype">void</span> VisitDequantizeLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer,</div><div class="line"><a name="l02792"></a><span class="lineno"> 2792</span>&#160;                              <span class="keyword">const</span> <span class="keywordtype">char</span>* name = <span class="keyword">nullptr</span>)<span class="keyword"> override</span></div><div class="line"><a name="l02793"></a><span class="lineno"> 2793</span>&#160;<span class="keyword">    </span>{</div><div class="line"><a name="l02794"></a><span class="lineno"> 2794</span>&#160;        boost::ignore_unused(layer, name);</div><div class="line"><a name="l02795"></a><span class="lineno"> 2795</span>&#160;        m_VisitedDequantizeLayer = <span class="keyword">true</span>;</div><div class="line"><a name="l02796"></a><span class="lineno"> 2796</span>&#160;    }</div><div class="line"><a name="l02797"></a><span class="lineno"> 2797</span>&#160;</div><div class="line"><a name="l02798"></a><span class="lineno"> 2798</span>&#160;    <span class="keywordtype">void</span> CheckQuantizeDequantizeLayerVisited(<span class="keywordtype">bool</span> expected)</div><div class="line"><a name="l02799"></a><span class="lineno"> 2799</span>&#160;    {</div><div class="line"><a name="l02800"></a><span class="lineno"> 2800</span>&#160;        <span class="keywordflow">if</span> (expected)</div><div class="line"><a name="l02801"></a><span class="lineno"> 2801</span>&#160;        {</div><div class="line"><a name="l02802"></a><span class="lineno"> 2802</span>&#160;            <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a0c262ba6f6c189a2d092d127c1b7627b">BOOST_CHECK</a>(m_VisitedQuantizeLayer);</div><div class="line"><a name="l02803"></a><span class="lineno"> 2803</span>&#160;            <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a0c262ba6f6c189a2d092d127c1b7627b">BOOST_CHECK</a>(m_VisitedDequantizeLayer);</div><div class="line"><a name="l02804"></a><span class="lineno"> 2804</span>&#160;        }</div><div class="line"><a name="l02805"></a><span class="lineno"> 2805</span>&#160;        <span class="keywordflow">else</span></div><div class="line"><a name="l02806"></a><span class="lineno"> 2806</span>&#160;        {</div><div class="line"><a name="l02807"></a><span class="lineno"> 2807</span>&#160;            <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a0c262ba6f6c189a2d092d127c1b7627b">BOOST_CHECK</a>(!m_VisitedQuantizeLayer);</div><div class="line"><a name="l02808"></a><span class="lineno"> 2808</span>&#160;            <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a0c262ba6f6c189a2d092d127c1b7627b">BOOST_CHECK</a>(!m_VisitedDequantizeLayer);</div><div class="line"><a name="l02809"></a><span class="lineno"> 2809</span>&#160;        }</div><div class="line"><a name="l02810"></a><span class="lineno"> 2810</span>&#160;    }</div><div class="line"><a name="l02811"></a><span class="lineno"> 2811</span>&#160;<span class="keyword">private</span>:</div><div class="line"><a name="l02812"></a><span class="lineno"> 2812</span>&#160;    <span class="keyword">const</span> <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6">DataType</a> m_DataType;</div><div class="line"><a name="l02813"></a><span class="lineno"> 2813</span>&#160;    <span class="keywordtype">bool</span> m_VisitedQuantizeLayer;</div><div class="line"><a name="l02814"></a><span class="lineno"> 2814</span>&#160;    <span class="keywordtype">bool</span> m_VisitedDequantizeLayer;</div><div class="line"><a name="l02815"></a><span class="lineno"> 2815</span>&#160;};</div><div class="line"><a name="l02816"></a><span class="lineno"> 2816</span>&#160;</div><div class="line"><a name="l02817"></a><span class="lineno"><a class="line" href="namespacearmnn.html#abe34cf42d7c8515ecd15d11f4aeb399c"> 2817</a></span>&#160;<span class="keywordtype">void</span> <a class="code" href="namespacearmnn.html#abe34cf42d7c8515ecd15d11f4aeb399c">PreserveTypeTestImpl</a>(<span class="keyword">const</span> <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6">DataType</a>&amp; dataType)</div><div class="line"><a name="l02818"></a><span class="lineno"> 2818</span>&#160;{</div><div class="line"><a name="l02819"></a><span class="lineno"> 2819</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l02820"></a><span class="lineno"> 2820</span>&#160;</div><div class="line"><a name="l02821"></a><span class="lineno"> 2821</span>&#160;    <span class="comment">// Add the layers</span></div><div class="line"><a name="l02822"></a><span class="lineno"> 2822</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input0 = network-&gt;AddInputLayer(0);</div><div class="line"><a name="l02823"></a><span class="lineno"> 2823</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* input1 = network-&gt;AddInputLayer(1);</div><div class="line"><a name="l02824"></a><span class="lineno"> 2824</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* addition = network-&gt;AddAdditionLayer();</div><div class="line"><a name="l02825"></a><span class="lineno"> 2825</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* output = network-&gt;AddOutputLayer(2);</div><div class="line"><a name="l02826"></a><span class="lineno"> 2826</span>&#160;</div><div class="line"><a name="l02827"></a><span class="lineno"> 2827</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02828"></a><span class="lineno"> 2828</span>&#160;    input1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(1));</div><div class="line"><a name="l02829"></a><span class="lineno"> 2829</span>&#160;    addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(output-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02830"></a><span class="lineno"> 2830</span>&#160;</div><div class="line"><a name="l02831"></a><span class="lineno"> 2831</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a> shape{1U, 2U, 3U};</div><div class="line"><a name="l02832"></a><span class="lineno"> 2832</span>&#160;    <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a> <a class="code" href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>(shape, dataType);</div><div class="line"><a name="l02833"></a><span class="lineno"> 2833</span>&#160;    input0-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l02834"></a><span class="lineno"> 2834</span>&#160;    input1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l02835"></a><span class="lineno"> 2835</span>&#160;    addition-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(info);</div><div class="line"><a name="l02836"></a><span class="lineno"> 2836</span>&#160;</div><div class="line"><a name="l02837"></a><span class="lineno"> 2837</span>&#160;    <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a> <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a> = dataType == <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a> ?</div><div class="line"><a name="l02838"></a><span class="lineno"> 2838</span>&#160;            <a class="code" href="structarmnn_1_1_quantizer_options.html">QuantizerOptions</a>(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a0a3f57c876f5a230244c38e1453a8a6e">DataType::QAsymmU8</a>, <span class="keyword">true</span>) : QuantizerOptions(dataType, <span class="keyword">true</span>);</div><div class="line"><a name="l02839"></a><span class="lineno"> 2839</span>&#160;</div><div class="line"><a name="l02840"></a><span class="lineno"> 2840</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantizedNetworkQAsymmU8 = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">INetworkQuantizer::Create</a>(network.get(), <a class="code" href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a>)-&gt;ExportNetwork();</div><div class="line"><a name="l02841"></a><span class="lineno"> 2841</span>&#160;    TestPreserveType validatorQAsymmU8(options, dataType, shape, shape);</div><div class="line"><a name="l02842"></a><span class="lineno"> 2842</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantizedNetworkQAsymmU8.get(), validatorQAsymmU8);</div><div class="line"><a name="l02843"></a><span class="lineno"> 2843</span>&#160;    validatorQAsymmU8.CheckQuantizeDequantizeLayerVisited(</div><div class="line"><a name="l02844"></a><span class="lineno"> 2844</span>&#160;    dataType == <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a> || dataType == <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a26e6ed77470c6f2f830ecf874e6c0d55">DataType::Float16</a>);</div><div class="line"><a name="l02845"></a><span class="lineno"> 2845</span>&#160;}</div><div class="line"><a name="l02846"></a><span class="lineno"> 2846</span>&#160;</div><div class="line"><a name="l02847"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a94eb3bdf0e1c8c748c2e29dce048ace4"> 2847</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(PreserveTypeFloat32)</div><div class="line"><a name="l02848"></a><span class="lineno"> 2848</span>&#160;{</div><div class="line"><a name="l02849"></a><span class="lineno"> 2849</span>&#160;    <a class="code" href="namespacearmnn.html#abe34cf42d7c8515ecd15d11f4aeb399c">PreserveTypeTestImpl</a>(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>);</div><div class="line"><a name="l02850"></a><span class="lineno"> 2850</span>&#160;}</div><div class="line"><a name="l02851"></a><span class="lineno"> 2851</span>&#160;</div><div class="line"><a name="l02852"></a><span class="lineno"><a class="line" href="namespacearmnn.html#ab242670b85e047e79bb297cdb192cc93"> 2852</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(PreserveTypeQAsymmU8)</div><div class="line"><a name="l02853"></a><span class="lineno"> 2853</span>&#160;{</div><div class="line"><a name="l02854"></a><span class="lineno"> 2854</span>&#160;    <a class="code" href="namespacearmnn.html#abe34cf42d7c8515ecd15d11f4aeb399c">PreserveTypeTestImpl</a>(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a0a3f57c876f5a230244c38e1453a8a6e">DataType::QAsymmU8</a>);</div><div class="line"><a name="l02855"></a><span class="lineno"> 2855</span>&#160;}</div><div class="line"><a name="l02856"></a><span class="lineno"> 2856</span>&#160;</div><div class="line"><a name="l02857"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a061891029598224370aae4cd18b78406"> 2857</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(PreserveTypeQsymm8)</div><div class="line"><a name="l02858"></a><span class="lineno"> 2858</span>&#160;{</div><div class="line"><a name="l02859"></a><span class="lineno"> 2859</span>&#160;    <a class="code" href="namespacearmnn.html#abe34cf42d7c8515ecd15d11f4aeb399c">PreserveTypeTestImpl</a>(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">DataType::QSymmS8</a>);</div><div class="line"><a name="l02860"></a><span class="lineno"> 2860</span>&#160;}</div><div class="line"><a name="l02861"></a><span class="lineno"> 2861</span>&#160;</div><div class="line"><a name="l02862"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a4d4386cbb19dbc551e423992ecdd0d61"> 2862</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(PreserveTypeQsymm16)</div><div class="line"><a name="l02863"></a><span class="lineno"> 2863</span>&#160;{</div><div class="line"><a name="l02864"></a><span class="lineno"> 2864</span>&#160;    <a class="code" href="namespacearmnn.html#abe34cf42d7c8515ecd15d11f4aeb399c">PreserveTypeTestImpl</a>(<a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">DataType::QSymmS16</a>);</div><div class="line"><a name="l02865"></a><span class="lineno"> 2865</span>&#160;}</div><div class="line"><a name="l02866"></a><span class="lineno"> 2866</span>&#160;</div><div class="line"><a name="l02867"></a><span class="lineno"><a class="line" href="namespacearmnn.html#a8c09fbb75d2c2dea48926a540fc5cce9"> 2867</a></span>&#160;<a class="code" href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">BOOST_AUTO_TEST_CASE</a>(TestConnectionPreservationAfterDynamicQuant)</div><div class="line"><a name="l02868"></a><span class="lineno"> 2868</span>&#160;{</div><div class="line"><a name="l02869"></a><span class="lineno"> 2869</span>&#160;    <span class="keyword">class </span>TestConnectionPreservation : <span class="keyword">public</span> <a class="code" href="classarmnn_1_1_layer_visitor_base.html">LayerVisitorBase</a>&lt;VisitorNoThrowPolicy&gt;</div><div class="line"><a name="l02870"></a><span class="lineno"> 2870</span>&#160;    {</div><div class="line"><a name="l02871"></a><span class="lineno"> 2871</span>&#160;    <span class="keyword">public</span>:</div><div class="line"><a name="l02872"></a><span class="lineno"> 2872</span>&#160;        TestConnectionPreservation(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_graph.html">Graph</a>&amp; graph)</div><div class="line"><a name="l02873"></a><span class="lineno"> 2873</span>&#160;            : <a class="code" href="classarmnn_1_1_layer_visitor_base.html">LayerVisitorBase&lt;VisitorNoThrowPolicy&gt;</a>()</div><div class="line"><a name="l02874"></a><span class="lineno"> 2874</span>&#160;            , m_Graph(graph)</div><div class="line"><a name="l02875"></a><span class="lineno"> 2875</span>&#160;        {}</div><div class="line"><a name="l02876"></a><span class="lineno"> 2876</span>&#160;</div><div class="line"><a name="l02877"></a><span class="lineno"> 2877</span>&#160;        <span class="keywordtype">void</span> VisitAdditionLayer(<span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* layer, <span class="keyword">const</span> <span class="keywordtype">char</span>*)<span class="keyword"> override</span></div><div class="line"><a name="l02878"></a><span class="lineno"> 2878</span>&#160;<span class="keyword">        </span>{</div><div class="line"><a name="l02879"></a><span class="lineno"> 2879</span>&#160;            CheckLayerName(layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_input_slot.html#a81fbf6103761e55061b62ba989b00f10">GetConnection</a>()-&gt;<a class="code" href="classarmnn_1_1_i_output_slot.html#ad0c3555b126975ad6b3e250fe2a59534">GetOwningLayerGuid</a>(), <span class="stringliteral">&quot;reLU1&quot;</span>);</div><div class="line"><a name="l02880"></a><span class="lineno"> 2880</span>&#160;            CheckLayerName(layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(1).<a class="code" href="classarmnn_1_1_i_input_slot.html#a81fbf6103761e55061b62ba989b00f10">GetConnection</a>()-&gt;<a class="code" href="classarmnn_1_1_i_output_slot.html#ad0c3555b126975ad6b3e250fe2a59534">GetOwningLayerGuid</a>(), <span class="stringliteral">&quot;reLU2&quot;</span>);</div><div class="line"><a name="l02881"></a><span class="lineno"> 2881</span>&#160;        }</div><div class="line"><a name="l02882"></a><span class="lineno"> 2882</span>&#160;</div><div class="line"><a name="l02883"></a><span class="lineno"> 2883</span>&#160;        <span class="keywordtype">void</span> CheckLayerName(<a class="code" href="classarmnn_1_1profiling_1_1_profiling_guid.html">LayerGuid</a> guid, std::string expectedName)</div><div class="line"><a name="l02884"></a><span class="lineno"> 2884</span>&#160;        {</div><div class="line"><a name="l02885"></a><span class="lineno"> 2885</span>&#160;            <span class="keywordtype">bool</span> guidFound = <span class="keyword">false</span>;</div><div class="line"><a name="l02886"></a><span class="lineno"> 2886</span>&#160;            <span class="keywordflow">for</span> (<a class="code" href="classarmnn_1_1_layer.html">Layer</a>* layer : m_Graph)</div><div class="line"><a name="l02887"></a><span class="lineno"> 2887</span>&#160;            {</div><div class="line"><a name="l02888"></a><span class="lineno"> 2888</span>&#160;                <span class="keywordflow">if</span> (layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#afb5e65c770f6cee222db8af7581541a6">GetGuid</a>() == guid)</div><div class="line"><a name="l02889"></a><span class="lineno"> 2889</span>&#160;                {</div><div class="line"><a name="l02890"></a><span class="lineno"> 2890</span>&#160;                    BOOST_CHECK_EQUAL(layer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#afcc1c3a20bd2860e0ddd21674389246f">GetName</a>(), expectedName.c_str());</div><div class="line"><a name="l02891"></a><span class="lineno"> 2891</span>&#160;                    guidFound = <span class="keyword">true</span>;</div><div class="line"><a name="l02892"></a><span class="lineno"> 2892</span>&#160;                    <span class="keywordflow">break</span>;</div><div class="line"><a name="l02893"></a><span class="lineno"> 2893</span>&#160;                }</div><div class="line"><a name="l02894"></a><span class="lineno"> 2894</span>&#160;            }</div><div class="line"><a name="l02895"></a><span class="lineno"> 2895</span>&#160;            <span class="keywordflow">if</span> (!guidFound)</div><div class="line"><a name="l02896"></a><span class="lineno"> 2896</span>&#160;            {</div><div class="line"><a name="l02897"></a><span class="lineno"> 2897</span>&#160;                BOOST_FAIL(<span class="stringliteral">&quot;No layer matching the GUID was found&quot;</span>);</div><div class="line"><a name="l02898"></a><span class="lineno"> 2898</span>&#160;            }</div><div class="line"><a name="l02899"></a><span class="lineno"> 2899</span>&#160;        }</div><div class="line"><a name="l02900"></a><span class="lineno"> 2900</span>&#160;</div><div class="line"><a name="l02901"></a><span class="lineno"> 2901</span>&#160;    <span class="keyword">private</span>:</div><div class="line"><a name="l02902"></a><span class="lineno"> 2902</span>&#160;        <a class="code" href="classarmnn_1_1_graph.html">Graph</a> m_Graph;</div><div class="line"><a name="l02903"></a><span class="lineno"> 2903</span>&#160;    };</div><div class="line"><a name="l02904"></a><span class="lineno"> 2904</span>&#160;</div><div class="line"><a name="l02905"></a><span class="lineno"> 2905</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> network = <a class="code" href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">INetwork::Create</a>();</div><div class="line"><a name="l02906"></a><span class="lineno"> 2906</span>&#160;</div><div class="line"><a name="l02907"></a><span class="lineno"> 2907</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* inputLayer =  network-&gt;AddInputLayer(0,<span class="stringliteral">&quot;inputLayer1&quot;</span>);</div><div class="line"><a name="l02908"></a><span class="lineno"> 2908</span>&#160;    <a class="code" href="structarmnn_1_1_activation_descriptor.html">armnn::ActivationDescriptor</a> ReLUDesc;</div><div class="line"><a name="l02909"></a><span class="lineno"> 2909</span>&#160;    ReLUDesc.<a class="code" href="structarmnn_1_1_activation_descriptor.html#af10fa7883e3579950f477bee92a64844">m_Function</a> = <a class="code" href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaa3d90c0a5ab3fcf8e6f6608cb3d3a1559">ActivationFunction::ReLu</a>;</div><div class="line"><a name="l02910"></a><span class="lineno"> 2910</span>&#160;</div><div class="line"><a name="l02911"></a><span class="lineno"> 2911</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* reLULayer1 = network-&gt;AddActivationLayer(ReLUDesc, <span class="stringliteral">&quot;reLU1&quot;</span>);</div><div class="line"><a name="l02912"></a><span class="lineno"> 2912</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* reLULayer2 = network-&gt;AddActivationLayer(ReLUDesc, <span class="stringliteral">&quot;reLU2&quot;</span>);</div><div class="line"><a name="l02913"></a><span class="lineno"> 2913</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* addLayer1 = network-&gt;AddAdditionLayer(<span class="stringliteral">&quot;addLayer1&quot;</span>);</div><div class="line"><a name="l02914"></a><span class="lineno"> 2914</span>&#160;    <a class="code" href="classarmnn_1_1_i_connectable_layer.html">IConnectableLayer</a>* outputLayer = network-&gt;AddOutputLayer(0,<span class="stringliteral">&quot;outPutLayer1&quot;</span>);</div><div class="line"><a name="l02915"></a><span class="lineno"> 2915</span>&#160;</div><div class="line"><a name="l02916"></a><span class="lineno"> 2916</span>&#160;    inputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(reLULayer1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02917"></a><span class="lineno"> 2917</span>&#160;    reLULayer1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(reLULayer2-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02918"></a><span class="lineno"> 2918</span>&#160;    reLULayer1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(addLayer1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02919"></a><span class="lineno"> 2919</span>&#160;    reLULayer2-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(addLayer1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(1));</div><div class="line"><a name="l02920"></a><span class="lineno"> 2920</span>&#160;    addLayer1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">Connect</a>(outputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">GetInputSlot</a>(0));</div><div class="line"><a name="l02921"></a><span class="lineno"> 2921</span>&#160;</div><div class="line"><a name="l02922"></a><span class="lineno"> 2922</span>&#160;    inputLayer-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(<a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>(<a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>({1, 2, 2, 1}), <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>));</div><div class="line"><a name="l02923"></a><span class="lineno"> 2923</span>&#160;    reLULayer1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(<a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>(<a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>({1, 2, 2, 1}), <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>));</div><div class="line"><a name="l02924"></a><span class="lineno"> 2924</span>&#160;    reLULayer2-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(<a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>(<a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>({1, 2, 2, 1}), <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>));</div><div class="line"><a name="l02925"></a><span class="lineno"> 2925</span>&#160;    addLayer1-&gt;<a class="code" href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">GetOutputSlot</a>(0).<a class="code" href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">SetTensorInfo</a>(<a class="code" href="classarmnn_1_1_tensor_info.html">TensorInfo</a>(<a class="code" href="classarmnn_1_1_tensor_shape.html">TensorShape</a>({1, 2, 2, 1}), <a class="code" href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">DataType::Float32</a>));</div><div class="line"><a name="l02926"></a><span class="lineno"> 2926</span>&#160;</div><div class="line"><a name="l02927"></a><span class="lineno"> 2927</span>&#160;    TestConnectionPreservation visitor1(boost::polymorphic_downcast&lt;const Network*&gt;(network.get())-&gt;GetGraph());</div><div class="line"><a name="l02928"></a><span class="lineno"> 2928</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(network.get(), visitor1);</div><div class="line"><a name="l02929"></a><span class="lineno"> 2929</span>&#160;</div><div class="line"><a name="l02930"></a><span class="lineno"> 2930</span>&#160;    <a class="code" href="namespacearmnn.html#a41119e261eec9343888d2ceab1e4999a">armnn::INetworkQuantizerPtr</a> quantizer = <a class="code" href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">armnn::INetworkQuantizer::Create</a>(network.get());</div><div class="line"><a name="l02931"></a><span class="lineno"> 2931</span>&#160;</div><div class="line"><a name="l02932"></a><span class="lineno"> 2932</span>&#160;    <a class="code" href="classarmnn_1_1_tensor_info.html">armnn::TensorInfo</a> tensorInfo = <a class="code" href="namespacearmnn.html#ae52296dff1f4879854f320d59f92574e">GetInputTensorInfo</a>(boost::polymorphic_downcast&lt;const Network*&gt;(network.get()));</div><div class="line"><a name="l02933"></a><span class="lineno"> 2933</span>&#160;</div><div class="line"><a name="l02934"></a><span class="lineno"> 2934</span>&#160;    std::vector&lt;float&gt; inputData({0, 2, 0, 4});</div><div class="line"><a name="l02935"></a><span class="lineno"> 2935</span>&#160;    <a class="code" href="classarmnn_1_1_const_tensor.html">armnn::ConstTensor</a> inputTensor(tensorInfo, inputData.data());</div><div class="line"><a name="l02936"></a><span class="lineno"> 2936</span>&#160;</div><div class="line"><a name="l02937"></a><span class="lineno"> 2937</span>&#160;    <a class="code" href="namespacearmnn.html#aa01bce88f89975a5a031db4cc8861527">InputTensors</a> inputTensors;</div><div class="line"><a name="l02938"></a><span class="lineno"> 2938</span>&#160;    inputTensors.push_back(std::make_pair(0, inputTensor));</div><div class="line"><a name="l02939"></a><span class="lineno"> 2939</span>&#160;    quantizer-&gt;Refine(inputTensors);</div><div class="line"><a name="l02940"></a><span class="lineno"> 2940</span>&#160;</div><div class="line"><a name="l02941"></a><span class="lineno"> 2941</span>&#160;    <a class="code" href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">INetworkPtr</a> quantNetwork = quantizer-&gt;ExportNetwork();</div><div class="line"><a name="l02942"></a><span class="lineno"> 2942</span>&#160;</div><div class="line"><a name="l02943"></a><span class="lineno"> 2943</span>&#160;    TestConnectionPreservation visitor2(boost::polymorphic_downcast&lt;const Network*&gt;(quantNetwork.get())-&gt;GetGraph());</div><div class="line"><a name="l02944"></a><span class="lineno"> 2944</span>&#160;    <a class="code" href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">VisitLayersTopologically</a>(quantNetwork.get(), visitor2);</div><div class="line"><a name="l02945"></a><span class="lineno"> 2945</span>&#160;}</div><div class="line"><a name="l02946"></a><span class="lineno"> 2946</span>&#160;</div><div class="line"><a name="l02947"></a><span class="lineno"> 2947</span>&#160;<a class="code" href="_profiler_tests_8cpp.html#af7f71af5c6c124222dd1c42c5df892f4">BOOST_AUTO_TEST_SUITE_END</a>()</div><div class="line"><a name="l02948"></a><span class="lineno"> 2948</span>&#160;} <span class="comment">// namespace armnn</span></div><div class="ttc" id="structarmnn_1_1_activation_descriptor_html_a017b2990003a014234f13e999dc7c689"><div class="ttname"><a href="structarmnn_1_1_activation_descriptor.html#a017b2990003a014234f13e999dc7c689">armnn::ActivationDescriptor::m_A</a></div><div class="ttdeci">float m_A</div><div class="ttdoc">Alpha upper bound value used by the activation functions. (BoundedReLu, Linear, TanH). </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00037">Descriptors.hpp:37</a></div></div>
  88 <div class="ttc" id="classarmnn_1_1_i_input_slot_html_a81fbf6103761e55061b62ba989b00f10"><div class="ttname"><a href="classarmnn_1_1_i_input_slot.html#a81fbf6103761e55061b62ba989b00f10">armnn::IInputSlot::GetConnection</a></div><div class="ttdeci">virtual const IOutputSlot * GetConnection() const =0</div></div>
  89 <div class="ttc" id="namespacearmnn_html_ae7e8cbf71db6a490789ca6dcaa8deeaea6a061313d22e51e0f25b7cd4dc065233"><div class="ttname"><a href="namespacearmnn.html#ae7e8cbf71db6a490789ca6dcaa8deeaea6a061313d22e51e0f25b7cd4dc065233">armnn::ArgMinMaxFunction::Max</a></div></div>
  90 <div class="ttc" id="namespacearmnn_html_aad4b8cb9a4d882a48bc21510f0d1a938"><div class="ttname"><a href="namespacearmnn.html#aad4b8cb9a4d882a48bc21510f0d1a938">armnn::CreateNetworkWithFullyConnectedLayer</a></div><div class="ttdeci">INetworkPtr CreateNetworkWithFullyConnectedLayer(const bool biasEnabled, const TensorShape &amp;inputShape, const TensorShape &amp;outputShape)</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l00951">QuantizerTest.cpp:951</a></div></div>
  91 <div class="ttc" id="namespacearmnn_html_a1465480794787d2278d3f0d2e6d887b4"><div class="ttname"><a href="namespacearmnn.html#a1465480794787d2278d3f0d2e6d887b4">armnn::g_SymmS16QuantizationBase</a></div><div class="ttdeci">const float g_SymmS16QuantizationBase</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l00037">QuantizerTest.cpp:37</a></div></div>
  92 <div class="ttc" id="namespacearmnn_html_a56297e0f7b215eea46c818cb7528d9eaacb7667f5ec2f6e8a5943b781ba6c2735"><div class="ttname"><a href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaacb7667f5ec2f6e8a5943b781ba6c2735">armnn::ActivationFunction::LeakyReLu</a></div></div>
  93 <div class="ttc" id="classarmnn_1_1_i_connectable_layer_html_afb5e65c770f6cee222db8af7581541a6"><div class="ttname"><a href="classarmnn_1_1_i_connectable_layer.html#afb5e65c770f6cee222db8af7581541a6">armnn::IConnectableLayer::GetGuid</a></div><div class="ttdeci">virtual LayerGuid GetGuid() const =0</div></div>
  94 <div class="ttc" id="namespacearmnn_html_acd7f8820d124166a38c95bc8ad38811b"><div class="ttname"><a href="namespacearmnn.html#acd7f8820d124166a38c95bc8ad38811b">armnn::g_SymmS8QuantizationBase</a></div><div class="ttdeci">const float g_SymmS8QuantizationBase</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l00036">QuantizerTest.cpp:36</a></div></div>
  95 <div class="ttc" id="classarmnn_1_1_override_input_range_visitor_html"><div class="ttname"><a href="classarmnn_1_1_override_input_range_visitor.html">armnn::OverrideInputRangeVisitor</a></div><div class="ttdoc">Visitor object for overriding the input range of the quantized input layers in a network. </div><div class="ttdef"><b>Definition:</b> <a href="_override_input_range_visitor_8hpp_source.html#l00018">OverrideInputRangeVisitor.hpp:18</a></div></div>
  96 <div class="ttc" id="namespacearmnn_html_a245661fc96c9c4a9b898e1d98c8c6962"><div class="ttname"><a href="namespacearmnn.html#a245661fc96c9c4a9b898e1d98c8c6962">armnn::ValidateFullyConnectedLayer</a></div><div class="ttdeci">void ValidateFullyConnectedLayer(const bool biasEnabled)</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l00989">QuantizerTest.cpp:989</a></div></div>
  97 <div class="ttc" id="classarmnn_1_1_i_connectable_layer_html_a6ec9e0eb66d7d6a01240492a0b18104c"><div class="ttname"><a href="classarmnn_1_1_i_connectable_layer.html#a6ec9e0eb66d7d6a01240492a0b18104c">armnn::IConnectableLayer::GetInputSlot</a></div><div class="ttdeci">virtual const IInputSlot &amp; GetInputSlot(unsigned int index) const =0</div></div>
  98 <div class="ttc" id="classarmnn_1_1_tensor_info_html_a770b51078da02f44a819e9f95d8058b5"><div class="ttname"><a href="classarmnn_1_1_tensor_info.html#a770b51078da02f44a819e9f95d8058b5">armnn::TensorInfo::GetQuantizationOffset</a></div><div class="ttdeci">int32_t GetQuantizationOffset() const</div><div class="ttdef"><b>Definition:</b> <a href="_tensor_8cpp_source.html#l00264">Tensor.cpp:264</a></div></div>
  99 <div class="ttc" id="namespacearmnn_html_ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db"><div class="ttname"><a href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9945327825b115e93a3b89f4302e76db">armnn::DataType::QSymmS8</a></div></div>
 100 <div class="ttc" id="structarmnn_1_1_batch_to_space_nd_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_batch_to_space_nd_descriptor.html">armnn::BatchToSpaceNdDescriptor</a></div><div class="ttdoc">A BatchToSpaceNdDescriptor for the BatchToSpaceNdLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00657">Descriptors.hpp:657</a></div></div>
 101 <div class="ttc" id="classarmnn_1_1_const_tensor_html"><div class="ttname"><a href="classarmnn_1_1_const_tensor.html">armnn::ConstTensor</a></div><div class="ttdoc">A tensor defined by a TensorInfo (shape and data type) and an immutable backing store. </div><div class="ttdef"><b>Definition:</b> <a href="_tensor_8hpp_source.html#l00199">Tensor.hpp:199</a></div></div>
 102 <div class="ttc" id="classarmnn_1_1_i_output_slot_html_a9943775a364fc4ab53b85ac88f311886"><div class="ttname"><a href="classarmnn_1_1_i_output_slot.html#a9943775a364fc4ab53b85ac88f311886">armnn::IOutputSlot::GetTensorInfo</a></div><div class="ttdeci">virtual const TensorInfo &amp; GetTensorInfo() const =0</div></div>
 103 <div class="ttc" id="namespacearmnn_html_abe34cf42d7c8515ecd15d11f4aeb399c"><div class="ttname"><a href="namespacearmnn.html#abe34cf42d7c8515ecd15d11f4aeb399c">armnn::PreserveTypeTestImpl</a></div><div class="ttdeci">void PreserveTypeTestImpl(const DataType &amp;dataType)</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l02817">QuantizerTest.cpp:2817</a></div></div>
 104 <div class="ttc" id="structarmnn_1_1_stand_in_descriptor_html_abb8a2d2bb8cc594c26aaa70c820ac5cc"><div class="ttname"><a href="structarmnn_1_1_stand_in_descriptor.html#abb8a2d2bb8cc594c26aaa70c820ac5cc">armnn::StandInDescriptor::m_NumOutputs</a></div><div class="ttdeci">uint32_t m_NumOutputs</div><div class="ttdoc">Number of output tensors. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00998">Descriptors.hpp:998</a></div></div>
 105 <div class="ttc" id="namespacearmnn_html_a56297e0f7b215eea46c818cb7528d9eaa3d90c0a5ab3fcf8e6f6608cb3d3a1559"><div class="ttname"><a href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaa3d90c0a5ab3fcf8e6f6608cb3d3a1559">armnn::ActivationFunction::ReLu</a></div></div>
 106 <div class="ttc" id="namespacearmnn_html_ad8ed01ff3ff33333d8e19db4d2818bb6a0a3f57c876f5a230244c38e1453a8a6e"><div class="ttname"><a href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a0a3f57c876f5a230244c38e1453a8a6e">armnn::DataType::QAsymmU8</a></div></div>
 107 <div class="ttc" id="classarmnn_1_1_network_html_a90d8841cfbbc82ab02328f33fed24ac6"><div class="ttname"><a href="classarmnn_1_1_network.html#a90d8841cfbbc82ab02328f33fed24ac6">armnn::Network::AddInputLayer</a></div><div class="ttdeci">IConnectableLayer * AddInputLayer(LayerBindingId id, const char *name=nullptr) override</div><div class="ttdef"><b>Definition:</b> <a href="_network_8cpp_source.html#l00953">Network.cpp:953</a></div></div>
 108 <div class="ttc" id="namespacearmnn_html_a9835ef753dda5b5a2fe827680e41fda7"><div class="ttname"><a href="namespacearmnn.html#a9835ef753dda5b5a2fe827680e41fda7">armnn::VisitLayers</a></div><div class="ttdeci">void VisitLayers(const LayerContainer &amp;layerContainer, ILayerVisitor &amp;visitor)</div><div class="ttdef"><b>Definition:</b> <a href="_network_quantizer_utils_8hpp_source.html#l00050">NetworkQuantizerUtils.hpp:50</a></div></div>
 109 <div class="ttc" id="namespacearmnn_html_ac757baefa4b72b54c38f713f86418f8a"><div class="ttname"><a href="namespacearmnn.html#ac757baefa4b72b54c38f713f86418f8a">armnn::MinMaxRanges</a></div><div class="ttdeci">std::vector&lt; MinMaxRange &gt; MinMaxRanges</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l00030">QuantizerTest.cpp:30</a></div></div>
 110 <div class="ttc" id="classarmnn_1_1_optional_html"><div class="ttname"><a href="classarmnn_1_1_optional.html">armnn::Optional</a></div><div class="ttdef"><b>Definition:</b> <a href="_optional_8hpp_source.html#l00270">Optional.hpp:270</a></div></div>
 111 <div class="ttc" id="namespacearmnn_html_a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c"><div class="ttname"><a href="namespacearmnn.html#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">armnn::BoostLogSeverityMapping::info</a></div></div>
 112 <div class="ttc" id="structarmnn_1_1_resize_descriptor_html_a46c3fa15c46fb0d1dcdc24d0ea5cb5cd"><div class="ttname"><a href="structarmnn_1_1_resize_descriptor.html#a46c3fa15c46fb0d1dcdc24d0ea5cb5cd">armnn::ResizeDescriptor::m_TargetHeight</a></div><div class="ttdeci">uint32_t m_TargetHeight</div><div class="ttdoc">Target height value. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00746">Descriptors.hpp:746</a></div></div>
 113 <div class="ttc" id="namespacearmnn_html_ae52296dff1f4879854f320d59f92574e"><div class="ttname"><a href="namespacearmnn.html#ae52296dff1f4879854f320d59f92574e">armnn::GetInputTensorInfo</a></div><div class="ttdeci">TensorInfo GetInputTensorInfo(const Network *network)</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l00337">QuantizerTest.cpp:337</a></div></div>
 114 <div class="ttc" id="classarmnn_1_1_layer_html"><div class="ttname"><a href="classarmnn_1_1_layer.html">armnn::Layer</a></div><div class="ttdef"><b>Definition:</b> <a href="_layer_8hpp_source.html#l00209">Layer.hpp:209</a></div></div>
 115 <div class="ttc" id="classarmnn_1_1_tensor_info_html"><div class="ttname"><a href="classarmnn_1_1_tensor_info.html">armnn::TensorInfo</a></div><div class="ttdef"><b>Definition:</b> <a href="_tensor_8hpp_source.html#l00053">Tensor.hpp:53</a></div></div>
 116 <div class="ttc" id="namespacearmnn_html_a10d15f3df1ab52b3b915a4be1dbf386b"><div class="ttname"><a href="namespacearmnn.html#a10d15f3df1ab52b3b915a4be1dbf386b">armnn::BOOST_AUTO_TEST_CASE</a></div><div class="ttdeci">BOOST_AUTO_TEST_CASE(CheckConvolution2dLayer)</div><div class="ttdef"><b>Definition:</b> <a href="_const_tensor_layer_visitor_8cpp_source.html#l00170">ConstTensorLayerVisitor.cpp:170</a></div></div>
 117 <div class="ttc" id="structarmnn_1_1_activation_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_activation_descriptor.html">armnn::ActivationDescriptor</a></div><div class="ttdoc">An ActivationDescriptor for the ActivationLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00020">Descriptors.hpp:20</a></div></div>
 118 <div class="ttc" id="namespacearmnn_html_a1a9a6dea47de10df8e3c76dd45df56fb"><div class="ttname"><a href="namespacearmnn.html#a1a9a6dea47de10df8e3c76dd45df56fb">armnn::g_TestTolerance</a></div><div class="ttdeci">const float g_TestTolerance</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l00038">QuantizerTest.cpp:38</a></div></div>
 119 <div class="ttc" id="classarmnn_1_1_range_tracker_html_a997e96288bdb106c922202e3f33d5d7b"><div class="ttname"><a href="classarmnn_1_1_range_tracker.html#a997e96288bdb106c922202e3f33d5d7b">armnn::RangeTracker::MinMaxRange</a></div><div class="ttdeci">std::pair&lt; float, float &gt; MinMaxRange</div><div class="ttdef"><b>Definition:</b> <a href="_range_tracker_8hpp_source.html#l00020">RangeTracker.hpp:20</a></div></div>
 120 <div class="ttc" id="structarmnn_1_1_batch_normalization_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_batch_normalization_descriptor.html">armnn::BatchNormalizationDescriptor</a></div><div class="ttdoc">A BatchNormalizationDescriptor for the BatchNormalizationLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00610">Descriptors.hpp:610</a></div></div>
 121 <div class="ttc" id="namespacearmnn_html_a56297e0f7b215eea46c818cb7528d9eaa1e34af023adeb7d5f484f8eb4b9826b6"><div class="ttname"><a href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaa1e34af023adeb7d5f484f8eb4b9826b6">armnn::ActivationFunction::Abs</a></div></div>
 122 <div class="ttc" id="namespacearmnn_html_a0e2bce68a1f7eff47ead4d9a2804eb91"><div class="ttname"><a href="namespacearmnn.html#a0e2bce68a1f7eff47ead4d9a2804eb91">armnn::QuantizeConstant</a></div><div class="ttdeci">void QuantizeConstant(const srcType *src, uint8_t *dst, size_t numElements, float &amp;scale, int &amp;offset)</div><div class="ttdef"><b>Definition:</b> <a href="_network_quantizer_utils_8hpp_source.html#l00023">NetworkQuantizerUtils.hpp:23</a></div></div>
 123 <div class="ttc" id="classarmnn_1_1_network_html_adb60c75544796e23d7abc1ce0476f6d9"><div class="ttname"><a href="classarmnn_1_1_network.html#adb60c75544796e23d7abc1ce0476f6d9">armnn::Network::AddAdditionLayer</a></div><div class="ttdeci">IConnectableLayer * AddAdditionLayer(const char *name=nullptr) override</div><div class="ttdef"><b>Definition:</b> <a href="_network_8cpp_source.html#l01212">Network.cpp:1212</a></div></div>
 124 <div class="ttc" id="classarmnn_1_1_range_tracker_html_a8e12342fc420701fbffd97025421575a"><div class="ttname"><a href="classarmnn_1_1_range_tracker.html#a8e12342fc420701fbffd97025421575a">armnn::RangeTracker::IsEmpty</a></div><div class="ttdeci">bool IsEmpty() const</div><div class="ttdoc">Query function to check that the RangeTracker is empty. </div><div class="ttdef"><b>Definition:</b> <a href="_range_tracker_8hpp_source.html#l00029">RangeTracker.hpp:29</a></div></div>
 125 <div class="ttc" id="structarmnn_1_1_fully_connected_descriptor_html_aea202e14d8874cefd9a0f778022b7e25"><div class="ttname"><a href="structarmnn_1_1_fully_connected_descriptor.html#aea202e14d8874cefd9a0f778022b7e25">armnn::FullyConnectedDescriptor::m_BiasEnabled</a></div><div class="ttdeci">bool m_BiasEnabled</div><div class="ttdoc">Enable/disable bias. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00386">Descriptors.hpp:386</a></div></div>
 126 <div class="ttc" id="structarmnn_1_1_views_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_views_descriptor.html">armnn::ViewsDescriptor</a></div><div class="ttdoc">A ViewsDescriptor for the SplitterLayer. Descriptor to configure the splitting process. Number of Views must be equal to the number of outputs, and their order must match - e.g. first view corresponds to the first output, second view to the second output, etc. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00190">Descriptors.hpp:190</a></div></div>
 127 <div class="ttc" id="namespacearmnn_html_ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204"><div class="ttname"><a href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">armnn::DataType::Float32</a></div></div>
 128 <div class="ttc" id="structarmnn_1_1_arg_min_max_descriptor_html_ab1ae6f520bb1a4da191a0ae907477f23"><div class="ttname"><a href="structarmnn_1_1_arg_min_max_descriptor.html#ab1ae6f520bb1a4da191a0ae907477f23">armnn::ArgMinMaxDescriptor::m_Function</a></div><div class="ttdeci">ArgMinMaxFunction m_Function</div><div class="ttdoc">Specify if the function is to find Min or Max. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00056">Descriptors.hpp:56</a></div></div>
 129 <div class="ttc" id="classarmnn_1_1_invalid_argument_exception_html"><div class="ttname"><a href="classarmnn_1_1_invalid_argument_exception.html">armnn::InvalidArgumentException</a></div><div class="ttdef"><b>Definition:</b> <a href="_exceptions_8hpp_source.html#l00080">Exceptions.hpp:80</a></div></div>
 130 <div class="ttc" id="namespacearmnn_html_a120c131df35d78b3a56cb0f07decaf35"><div class="ttname"><a href="namespacearmnn.html#a120c131df35d78b3a56cb0f07decaf35">armnn::CreateStartOfLeakyReluNetwork</a></div><div class="ttdeci">IConnectableLayer * CreateStartOfLeakyReluNetwork(INetwork *network, const TensorInfo &amp;info)</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l01474">QuantizerTest.cpp:1474</a></div></div>
 131 <div class="ttc" id="classarmnn_1_1_i_output_slot_html_ad0c3555b126975ad6b3e250fe2a59534"><div class="ttname"><a href="classarmnn_1_1_i_output_slot.html#ad0c3555b126975ad6b3e250fe2a59534">armnn::IOutputSlot::GetOwningLayerGuid</a></div><div class="ttdeci">virtual LayerGuid GetOwningLayerGuid() const =0</div></div>
 132 <div class="ttc" id="structarmnn_1_1_space_to_depth_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_space_to_depth_descriptor.html">armnn::SpaceToDepthDescriptor</a></div><div class="ttdoc">A SpaceToDepthDescriptor for the SpaceToDepthLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00810">Descriptors.hpp:810</a></div></div>
 133 <div class="ttc" id="classarmnn_1_1_i_connectable_layer_html_afcc1c3a20bd2860e0ddd21674389246f"><div class="ttname"><a href="classarmnn_1_1_i_connectable_layer.html#afcc1c3a20bd2860e0ddd21674389246f">armnn::IConnectableLayer::GetName</a></div><div class="ttdeci">virtual const char * GetName() const =0</div></div>
 134 <div class="ttc" id="namespacearmnn_html_a5abbe8a9ee003c1379a921dbe2745b81"><div class="ttname"><a href="namespacearmnn.html#a5abbe8a9ee003c1379a921dbe2745b81">armnn::TestQuantizeDepthwiseConvolution2d</a></div><div class="ttdeci">void TestQuantizeDepthwiseConvolution2d(bool useBiases)</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l01132">QuantizerTest.cpp:1132</a></div></div>
 135 <div class="ttc" id="structarmnn_1_1_softmax_descriptor_html_a8275d51ef9a584feb95726ea0522f6e5"><div class="ttname"><a href="structarmnn_1_1_softmax_descriptor.html#a8275d51ef9a584feb95726ea0522f6e5">armnn::SoftmaxDescriptor::m_Beta</a></div><div class="ttdeci">float m_Beta</div><div class="ttdoc">Exponentiation value. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00136">Descriptors.hpp:136</a></div></div>
 136 <div class="ttc" id="classarmnn_1_1_i_output_slot_html_a5ee4a6c9a2481245487b1b1a70d20fd0"><div class="ttname"><a href="classarmnn_1_1_i_output_slot.html#a5ee4a6c9a2481245487b1b1a70d20fd0">armnn::IOutputSlot::SetTensorInfo</a></div><div class="ttdeci">virtual void SetTensorInfo(const TensorInfo &amp;tensorInfo)=0</div></div>
 137 <div class="ttc" id="structarmnn_1_1_reshape_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_reshape_descriptor.html">armnn::ReshapeDescriptor</a></div><div class="ttdoc">A ReshapeDescriptor for the ReshapeLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00758">Descriptors.hpp:758</a></div></div>
 138 <div class="ttc" id="structarmnn_1_1_transpose_convolution2d_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_transpose_convolution2d_descriptor.html">armnn::TransposeConvolution2dDescriptor</a></div><div class="ttdoc">A TransposeConvolution2dDescriptor for the TransposeConvolution2dLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l01079">Descriptors.hpp:1079</a></div></div>
 139 <div class="ttc" id="namespacearmnn_html_a997e96288bdb106c922202e3f33d5d7b"><div class="ttname"><a href="namespacearmnn.html#a997e96288bdb106c922202e3f33d5d7b">armnn::MinMaxRange</a></div><div class="ttdeci">std::pair&lt; float, float &gt; MinMaxRange</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l00029">QuantizerTest.cpp:29</a></div></div>
 140 <div class="ttc" id="classarmnn_1_1_test_output_layer_visitor_html"><div class="ttname"><a href="classarmnn_1_1_test_output_layer_visitor.html">armnn::TestOutputLayerVisitor</a></div><div class="ttdef"><b>Definition:</b> <a href="_test_input_output_layer_visitor_8hpp_source.html#l00040">TestInputOutputLayerVisitor.hpp:40</a></div></div>
 141 <div class="ttc" id="namespacearmnn_html_ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d"><div class="ttname"><a href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a9d02ea768c081d4bdb2b7cab0b3f510d">armnn::DataType::QAsymmS8</a></div></div>
 142 <div class="ttc" id="classarmnn_1_1_i_network_html_a87d5ec72def73ca14bd2987a024bd569"><div class="ttname"><a href="classarmnn_1_1_i_network.html#a87d5ec72def73ca14bd2987a024bd569">armnn::INetwork::AddInputLayer</a></div><div class="ttdeci">virtual IConnectableLayer * AddInputLayer(LayerBindingId id, const char *name=nullptr)=0</div></div>
 143 <div class="ttc" id="classarmnn_1_1_range_tracker_html_a084c5aacd7e3bb07f2cfd5a8e9b0dd30"><div class="ttname"><a href="classarmnn_1_1_range_tracker.html#a084c5aacd7e3bb07f2cfd5a8e9b0dd30">armnn::RangeTracker::HasRanges</a></div><div class="ttdeci">bool HasRanges(LayerGuid guid) const</div><div class="ttdoc">Query that there is an entry for a layer. </div><div class="ttdef"><b>Definition:</b> <a href="_range_tracker_8hpp_source.html#l00032">RangeTracker.hpp:32</a></div></div>
 144 <div class="ttc" id="namespacearmnn_html_ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7"><div class="ttname"><a href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a053c769dcf82d66ef326c86980c02ba7">armnn::DataType::QSymmS16</a></div></div>
 145 <div class="ttc" id="classarmnn_1_1_network_html"><div class="ttname"><a href="classarmnn_1_1_network.html">armnn::Network</a></div><div class="ttdoc">Private implementation of INetwork. </div><div class="ttdef"><b>Definition:</b> <a href="_network_8hpp_source.html#l00027">Network.hpp:27</a></div></div>
 146 <div class="ttc" id="classarmnn_1_1_i_layer_visitor_html"><div class="ttname"><a href="classarmnn_1_1_i_layer_visitor.html">armnn::ILayerVisitor</a></div><div class="ttdef"><b>Definition:</b> <a href="_i_layer_visitor_8hpp_source.html#l00016">ILayerVisitor.hpp:16</a></div></div>
 147 <div class="ttc" id="classarmnn_1_1_i_network_html"><div class="ttname"><a href="classarmnn_1_1_i_network.html">armnn::INetwork</a></div><div class="ttdef"><b>Definition:</b> <a href="_i_network_8hpp_source.html#l00089">INetwork.hpp:89</a></div></div>
 148 <div class="ttc" id="structarmnn_1_1_activation_descriptor_html_a28c4c9cb15f6be3499abbc46b356060b"><div class="ttname"><a href="structarmnn_1_1_activation_descriptor.html#a28c4c9cb15f6be3499abbc46b356060b">armnn::ActivationDescriptor::m_B</a></div><div class="ttdeci">float m_B</div><div class="ttdoc">Beta lower bound value used by the activation functions. (BoundedReLu, Linear, TanH). </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00039">Descriptors.hpp:39</a></div></div>
 149 <div class="ttc" id="namespacearmnn_html_a41119e261eec9343888d2ceab1e4999a"><div class="ttname"><a href="namespacearmnn.html#a41119e261eec9343888d2ceab1e4999a">armnn::INetworkQuantizerPtr</a></div><div class="ttdeci">std::unique_ptr&lt; class INetworkQuantizer, void(*)(INetworkQuantizer *quantizer)&gt; INetworkQuantizerPtr</div><div class="ttdef"><b>Definition:</b> <a href="_i_network_quantizer_8hpp_source.html#l00029">INetworkQuantizer.hpp:29</a></div></div>
 150 <div class="ttc" id="namespacearmnn_html_a5fbc1479db5f4ff70a750cf02d58971b"><div class="ttname"><a href="namespacearmnn.html#a5fbc1479db5f4ff70a750cf02d58971b">armnn::CreateNetworkWithActivationLayer</a></div><div class="ttdeci">INetworkPtr CreateNetworkWithActivationLayer(const ActivationDescriptor &amp;descriptor, const TensorShape &amp;shape)</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l00297">QuantizerTest.cpp:297</a></div></div>
 151 <div class="ttc" id="namespacearmnn_html_a2d299363c9fc33334c571fa29ca4f58caa4cbdbb6070a5abb35fc95ecf1e22c14"><div class="ttname"><a href="namespacearmnn.html#a2d299363c9fc33334c571fa29ca4f58caa4cbdbb6070a5abb35fc95ecf1e22c14">armnn::ComparisonOperation::LessOrEqual</a></div></div>
 152 <div class="ttc" id="namespacearmnn_html_aa01bce88f89975a5a031db4cc8861527"><div class="ttname"><a href="namespacearmnn.html#aa01bce88f89975a5a031db4cc8861527">armnn::InputTensors</a></div><div class="ttdeci">std::vector&lt; std::pair&lt; LayerBindingId, class ConstTensor &gt; &gt; InputTensors</div><div class="ttdef"><b>Definition:</b> <a href="_tensor_8hpp_source.html#l00225">Tensor.hpp:225</a></div></div>
 153 <div class="ttc" id="namespacearmnn_html_a6482907b4c57873e197324f5cb66fd4d"><div class="ttname"><a href="namespacearmnn.html#a6482907b4c57873e197324f5cb66fd4d">armnn::VisitLayersTopologically</a></div><div class="ttdeci">void VisitLayersTopologically(const INetwork *inputNetwork, ILayerVisitor &amp;visitor)</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l00193">QuantizerTest.cpp:193</a></div></div>
 154 <div class="ttc" id="classarmnn_1_1_tensor_info_html_a685739c4eb65a580e075282cfe6787d6"><div class="ttname"><a href="classarmnn_1_1_tensor_info.html#a685739c4eb65a580e075282cfe6787d6">armnn::TensorInfo::SetQuantizationScale</a></div><div class="ttdeci">void SetQuantizationScale(float scale)</div><div class="ttdef"><b>Definition:</b> <a href="_tensor_8cpp_source.html#l00259">Tensor.cpp:259</a></div></div>
 155 <div class="ttc" id="namespacearmnn_html_aa9c6c1a7b5380a99a536f4740f87dd59"><div class="ttname"><a href="namespacearmnn.html#aa9c6c1a7b5380a99a536f4740f87dd59">armnn::CreateNetworkWithInputOutputLayers</a></div><div class="ttdeci">INetworkPtr CreateNetworkWithInputOutputLayers()</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l00318">QuantizerTest.cpp:318</a></div></div>
 156 <div class="ttc" id="classarmnn_1_1_i_network_html_a706f7345af3f18f4b16e226a672214c6"><div class="ttname"><a href="classarmnn_1_1_i_network.html#a706f7345af3f18f4b16e226a672214c6">armnn::INetwork::Create</a></div><div class="ttdeci">static INetworkPtr Create()</div><div class="ttdef"><b>Definition:</b> <a href="_network_8cpp_source.html#l00048">Network.cpp:48</a></div></div>
 157 <div class="ttc" id="structarmnn_1_1_instance_normalization_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_instance_normalization_descriptor.html">armnn::InstanceNormalizationDescriptor</a></div><div class="ttdoc">An InstanceNormalizationDescriptor for InstanceNormalizationLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00629">Descriptors.hpp:629</a></div></div>
 158 <div class="ttc" id="_file_only_profiling_decorator_tests_8cpp_html_a0c262ba6f6c189a2d092d127c1b7627b"><div class="ttname"><a href="_file_only_profiling_decorator_tests_8cpp.html#a0c262ba6f6c189a2d092d127c1b7627b">BOOST_CHECK</a></div><div class="ttdeci">BOOST_CHECK(profilingService.GetCurrentState()==ProfilingState::WaitingForAck)</div></div>
 159 <div class="ttc" id="structarmnn_1_1_fully_connected_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_fully_connected_descriptor.html">armnn::FullyConnectedDescriptor</a></div><div class="ttdoc">A FullyConnectedDescriptor for the FullyConnectedLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00373">Descriptors.hpp:373</a></div></div>
 160 <div class="ttc" id="_quantize_helper_8hpp_html"><div class="ttname"><a href="_quantize_helper_8hpp.html">QuantizeHelper.hpp</a></div></div>
 161 <div class="ttc" id="namespacearmnn_html"><div class="ttname"><a href="namespacearmnn.html">armnn</a></div><div class="ttdef"><b>Definition:</b> <a href="_backend_helper_8hpp_source.html#l00011">BackendHelper.hpp:11</a></div></div>
 162 <div class="ttc" id="_types_8hpp_html"><div class="ttname"><a href="_types_8hpp.html">Types.hpp</a></div></div>
 163 <div class="ttc" id="_layer_visitor_base_8hpp_html"><div class="ttname"><a href="_layer_visitor_base_8hpp.html">LayerVisitorBase.hpp</a></div></div>
 164 <div class="ttc" id="classarmnn_1_1_graph_html_a919fb58873ef3a6549e4490e226f2eae"><div class="ttname"><a href="classarmnn_1_1_graph.html#a919fb58873ef3a6549e4490e226f2eae">armnn::Graph::GetInputLayers</a></div><div class="ttdeci">InputLayersAccessor GetInputLayers() const</div><div class="ttdef"><b>Definition:</b> <a href="_graph_8hpp_source.html#l00181">Graph.hpp:181</a></div></div>
 165 <div class="ttc" id="namespacearmnn_html_a56297e0f7b215eea46c818cb7528d9eaa6bba7052636d1740303b1b2ef3b53fef"><div class="ttname"><a href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaa6bba7052636d1740303b1b2ef3b53fef">armnn::ActivationFunction::SoftReLu</a></div></div>
 166 <div class="ttc" id="namespacearmnn_html_ace74f6f9feb95a964a49d79458232703"><div class="ttname"><a href="namespacearmnn.html#ace74f6f9feb95a964a49d79458232703">armnn::INetworkPtr</a></div><div class="ttdeci">std::unique_ptr&lt; INetwork, void(*)(INetwork *network)&gt; INetworkPtr</div><div class="ttdef"><b>Definition:</b> <a href="_i_network_8hpp_source.html#l00085">INetwork.hpp:85</a></div></div>
 167 <div class="ttc" id="structarmnn_1_1_convolution2d_descriptor_html_aea202e14d8874cefd9a0f778022b7e25"><div class="ttname"><a href="structarmnn_1_1_convolution2d_descriptor.html#aea202e14d8874cefd9a0f778022b7e25">armnn::Convolution2dDescriptor::m_BiasEnabled</a></div><div class="ttdeci">bool m_BiasEnabled</div><div class="ttdoc">Enable/disable bias. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00438">Descriptors.hpp:438</a></div></div>
 168 <div class="ttc" id="namespacearmnn_html_ad8ed01ff3ff33333d8e19db4d2818bb6accedffbc6e5308e33d3843e8bdc0dad7"><div class="ttname"><a href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6accedffbc6e5308e33d3843e8bdc0dad7">armnn::DataType::Signed32</a></div></div>
 169 <div class="ttc" id="classarmnn_1_1_network_html_afe0a4f719f9752a405e71878da7012ba"><div class="ttname"><a href="classarmnn_1_1_network.html#afe0a4f719f9752a405e71878da7012ba">armnn::Network::GetGraph</a></div><div class="ttdeci">const Graph &amp; GetGraph() const</div><div class="ttdef"><b>Definition:</b> <a href="_network_8hpp_source.html#l00033">Network.hpp:33</a></div></div>
 170 <div class="ttc" id="namespacearmnn_html_a061aafb62b3769f55369845c3990ec7a"><div class="ttname"><a href="namespacearmnn.html#a061aafb62b3769f55369845c3990ec7a">armnn::MinMaxRangeMap</a></div><div class="ttdeci">std::unordered_map&lt; LayerGuid, MinMaxRanges &gt; MinMaxRangeMap</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l00031">QuantizerTest.cpp:31</a></div></div>
 171 <div class="ttc" id="classarmnn_1_1_base_tensor_html_a8aeddebdcf02e1832b22203c08a6b678"><div class="ttname"><a href="classarmnn_1_1_base_tensor.html#a8aeddebdcf02e1832b22203c08a6b678">armnn::BaseTensor::GetInfo</a></div><div class="ttdeci">const TensorInfo &amp; GetInfo() const</div><div class="ttdef"><b>Definition:</b> <a href="_tensor_8hpp_source.html#l00167">Tensor.hpp:167</a></div></div>
 172 <div class="ttc" id="structarmnn_1_1_stack_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_stack_descriptor.html">armnn::StackDescriptor</a></div><div class="ttdoc">A StackDescriptor for the StackLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00950">Descriptors.hpp:950</a></div></div>
 173 <div class="ttc" id="_profiler_tests_8cpp_html_af7f71af5c6c124222dd1c42c5df892f4"><div class="ttname"><a href="_profiler_tests_8cpp.html#af7f71af5c6c124222dd1c42c5df892f4">BOOST_AUTO_TEST_SUITE_END</a></div><div class="ttdeci">BOOST_AUTO_TEST_SUITE_END()</div></div>
 174 <div class="ttc" id="_i_network_quantizer_8hpp_html"><div class="ttname"><a href="_i_network_quantizer_8hpp.html">INetworkQuantizer.hpp</a></div></div>
 175 <div class="ttc" id="structarmnn_1_1_stand_in_descriptor_html_aed6086070440ceb94129bef06f70173f"><div class="ttname"><a href="structarmnn_1_1_stand_in_descriptor.html#aed6086070440ceb94129bef06f70173f">armnn::StandInDescriptor::m_NumInputs</a></div><div class="ttdeci">uint32_t m_NumInputs</div><div class="ttdoc">Number of input tensors. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00996">Descriptors.hpp:996</a></div></div>
 176 <div class="ttc" id="structarmnn_1_1_activation_descriptor_html_af10fa7883e3579950f477bee92a64844"><div class="ttname"><a href="structarmnn_1_1_activation_descriptor.html#af10fa7883e3579950f477bee92a64844">armnn::ActivationDescriptor::m_Function</a></div><div class="ttdeci">ActivationFunction m_Function</div><div class="ttdoc">The activation function to use (Sigmoid, TanH, Linear, ReLu, BoundedReLu, SoftReLu, LeakyReLu, Abs, Sqrt, Square). </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00035">Descriptors.hpp:35</a></div></div>
 177 <div class="ttc" id="structarmnn_1_1_softmax_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_softmax_descriptor.html">armnn::SoftmaxDescriptor</a></div><div class="ttdoc">A SoftmaxDescriptor for the SoftmaxLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00123">Descriptors.hpp:123</a></div></div>
 178 <div class="ttc" id="classarmnn_1_1_unimplemented_exception_html"><div class="ttname"><a href="classarmnn_1_1_unimplemented_exception.html">armnn::UnimplementedException</a></div><div class="ttdef"><b>Definition:</b> <a href="_exceptions_8hpp_source.html#l00098">Exceptions.hpp:98</a></div></div>
 179 <div class="ttc" id="namespacearmnn_html_a09bdfaa922d72ce0d9ec014dfa8f8c95"><div class="ttname"><a href="namespacearmnn.html#a09bdfaa922d72ce0d9ec014dfa8f8c95">armnn::g_AsymmS8QuantizationBase</a></div><div class="ttdeci">const float g_AsymmS8QuantizationBase</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l00035">QuantizerTest.cpp:35</a></div></div>
 180 <div class="ttc" id="structarmnn_1_1_depthwise_convolution2d_descriptor_html_aea202e14d8874cefd9a0f778022b7e25"><div class="ttname"><a href="structarmnn_1_1_depthwise_convolution2d_descriptor.html#aea202e14d8874cefd9a0f778022b7e25">armnn::DepthwiseConvolution2dDescriptor::m_BiasEnabled</a></div><div class="ttdeci">bool m_BiasEnabled</div><div class="ttdoc">Enable/disable bias. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00490">Descriptors.hpp:490</a></div></div>
 181 <div class="ttc" id="namespacearmnn_html_ad8ed01ff3ff33333d8e19db4d2818bb6"><div class="ttname"><a href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6">armnn::DataType</a></div><div class="ttdeci">DataType</div><div class="ttdef"><b>Definition:</b> <a href="_types_8hpp_source.html#l00032">Types.hpp:32</a></div></div>
 182 <div class="ttc" id="classarmnn_1_1_i_network_quantizer_html_a3a4d01d9351c02a703740290f226441f"><div class="ttname"><a href="classarmnn_1_1_i_network_quantizer.html#a3a4d01d9351c02a703740290f226441f">armnn::INetworkQuantizer::Create</a></div><div class="ttdeci">static INetworkQuantizerPtr Create(INetwork *inputNetwork, const QuantizerOptions &amp;options=QuantizerOptions())</div><div class="ttdoc">Create Quantizer object wrapped in unique_ptr. </div><div class="ttdef"><b>Definition:</b> <a href="_network_quantizer_8cpp_source.html#l00040">NetworkQuantizer.cpp:40</a></div></div>
 183 <div class="ttc" id="classarmnn_1_1_tensor_shape_html"><div class="ttname"><a href="classarmnn_1_1_tensor_shape.html">armnn::TensorShape</a></div><div class="ttdef"><b>Definition:</b> <a href="_tensor_8hpp_source.html#l00020">Tensor.hpp:20</a></div></div>
 184 <div class="ttc" id="classarmnn_1_1_range_tracker_html_a507bae23f59e94b4161886ebe663cdf4"><div class="ttname"><a href="classarmnn_1_1_range_tracker.html#a507bae23f59e94b4161886ebe663cdf4">armnn::RangeTracker::GetRange</a></div><div class="ttdeci">MinMaxRange GetRange(LayerGuid guid, unsigned int idx) const</div><div class="ttdoc">Retrieve the Range for a particular output slot on a particular layer. </div><div class="ttdef"><b>Definition:</b> <a href="_range_tracker_8cpp_source.html#l00029">RangeTracker.cpp:29</a></div></div>
 185 <div class="ttc" id="classarmnn_1_1_range_tracker_html"><div class="ttname"><a href="classarmnn_1_1_range_tracker.html">armnn::RangeTracker</a></div><div class="ttdef"><b>Definition:</b> <a href="_range_tracker_8hpp_source.html#l00017">RangeTracker.hpp:17</a></div></div>
 186 <div class="ttc" id="namespacearmnn_html_a56297e0f7b215eea46c818cb7528d9eaabc5a0f0d6e7cf7fca73299dcd46c0f0d"><div class="ttname"><a href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaabc5a0f0d6e7cf7fca73299dcd46c0f0d">armnn::ActivationFunction::BoundedReLu</a></div><div class="ttdoc">min(a, max(b, input)) </div></div>
 187 <div class="ttc" id="_tensor_8hpp_html"><div class="ttname"><a href="_tensor_8hpp.html">Tensor.hpp</a></div></div>
 188 <div class="ttc" id="classarmnn_1_1_i_connectable_layer_html"><div class="ttname"><a href="classarmnn_1_1_i_connectable_layer.html">armnn::IConnectableLayer</a></div><div class="ttdoc">Interface for a layer that is connectable to other layers via InputSlots and OutputSlots. </div><div class="ttdef"><b>Definition:</b> <a href="_i_network_8hpp_source.html#l00061">INetwork.hpp:61</a></div></div>
 189 <div class="ttc" id="namespacearmnn_html_ad8ed01ff3ff33333d8e19db4d2818bb6a26e6ed77470c6f2f830ecf874e6c0d55"><div class="ttname"><a href="namespacearmnn.html#ad8ed01ff3ff33333d8e19db4d2818bb6a26e6ed77470c6f2f830ecf874e6c0d55">armnn::DataType::Float16</a></div></div>
 190 <div class="ttc" id="classarmnn_1_1_graph_html"><div class="ttname"><a href="classarmnn_1_1_graph.html">armnn::Graph</a></div><div class="ttdef"><b>Definition:</b> <a href="_graph_8hpp_source.html#l00029">Graph.hpp:29</a></div></div>
 191 <div class="ttc" id="classarmnn_1_1_tensor_info_html_aea909c7327109228ef618d459015def3"><div class="ttname"><a href="classarmnn_1_1_tensor_info.html#aea909c7327109228ef618d459015def3">armnn::TensorInfo::GetDataType</a></div><div class="ttdeci">DataType GetDataType() const</div><div class="ttdef"><b>Definition:</b> <a href="_tensor_8hpp_source.html#l00095">Tensor.hpp:95</a></div></div>
 192 <div class="ttc" id="_output_shape_of_squeeze_8cpp_html_ae3a6cb217a792718f2bd0e8f45e3ca9e"><div class="ttname"><a href="_output_shape_of_squeeze_8cpp.html#ae3a6cb217a792718f2bd0e8f45e3ca9e">BOOST_AUTO_TEST_SUITE</a></div><div class="ttdeci">BOOST_AUTO_TEST_SUITE(TensorflowLiteParser)</div></div>
 193 <div class="ttc" id="namespacearmnn_html_a14cfd39cfc30682fa821ade3dd298426"><div class="ttname"><a href="namespacearmnn.html#a14cfd39cfc30682fa821ade3dd298426">armnn::TestQuantizeConvolution2d</a></div><div class="ttdeci">void TestQuantizeConvolution2d(bool useBiases)</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l01046">QuantizerTest.cpp:1046</a></div></div>
 194 <div class="ttc" id="namespacearmnn_html_a19994153bdbd7710f0af3973403bc4cc"><div class="ttname"><a href="namespacearmnn.html#a19994153bdbd7710f0af3973403bc4cc">armnn::g_AsymmU8QuantizationBase</a></div><div class="ttdeci">const float g_AsymmU8QuantizationBase</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l00033">QuantizerTest.cpp:33</a></div></div>
 195 <div class="ttc" id="structarmnn_1_1_pooling2d_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_pooling2d_descriptor.html">armnn::Pooling2dDescriptor</a></div><div class="ttdoc">A Pooling2dDescriptor for the Pooling2dLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00313">Descriptors.hpp:313</a></div></div>
 196 <div class="ttc" id="structarmnn_1_1_stand_in_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_stand_in_descriptor.html">armnn::StandInDescriptor</a></div><div class="ttdoc">A StandInDescriptor for the StandIn layer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00980">Descriptors.hpp:980</a></div></div>
 197 <div class="ttc" id="structarmnn_1_1_slice_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_slice_descriptor.html">armnn::SliceDescriptor</a></div><div class="ttdoc">A SliceDescriptor for the SliceLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00927">Descriptors.hpp:927</a></div></div>
 198 <div class="ttc" id="structarmnn_1_1_transpose_convolution2d_descriptor_html_aea202e14d8874cefd9a0f778022b7e25"><div class="ttname"><a href="structarmnn_1_1_transpose_convolution2d_descriptor.html#aea202e14d8874cefd9a0f778022b7e25">armnn::TransposeConvolution2dDescriptor::m_BiasEnabled</a></div><div class="ttdeci">bool m_BiasEnabled</div><div class="ttdoc">Enable/disable bias. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l01117">Descriptors.hpp:1117</a></div></div>
 199 <div class="ttc" id="classarmnn_1_1_i_network_html_aea068f6094e1c3bfcdf8167b68112632"><div class="ttname"><a href="classarmnn_1_1_i_network.html#aea068f6094e1c3bfcdf8167b68112632">armnn::INetwork::AddActivationLayer</a></div><div class="ttdeci">virtual IConnectableLayer * AddActivationLayer(const ActivationDescriptor &amp;activationDescriptor, const char *name=nullptr)=0</div></div>
 200 <div class="ttc" id="classarmnn_1_1_i_network_html_ad8582fba2ebeb65da43a56bc22d4f88b"><div class="ttname"><a href="classarmnn_1_1_i_network.html#ad8582fba2ebeb65da43a56bc22d4f88b">armnn::INetwork::AddOutputLayer</a></div><div class="ttdeci">virtual IConnectableLayer * AddOutputLayer(LayerBindingId id, const char *name=nullptr)=0</div></div>
 201 <div class="ttc" id="structarmnn_1_1_depthwise_convolution2d_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_depthwise_convolution2d_descriptor.html">armnn::DepthwiseConvolution2dDescriptor</a></div><div class="ttdoc">A DepthwiseConvolution2dDescriptor for the DepthwiseConvolution2dLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00444">Descriptors.hpp:444</a></div></div>
 202 <div class="ttc" id="namespacearmnn_html_a9b8e5a95f8c061bbbcdb036915dcb61a"><div class="ttname"><a href="namespacearmnn.html#a9b8e5a95f8c061bbbcdb036915dcb61a">armnn::OffsetScalePair</a></div><div class="ttdeci">std::pair&lt; float, int &gt; OffsetScalePair</div><div class="ttdef"><b>Definition:</b> <a href="_network_quantization_scheme_8hpp_source.html#l00016">NetworkQuantizationScheme.hpp:16</a></div></div>
 203 <div class="ttc" id="structarmnn_1_1_permute_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_permute_descriptor.html">armnn::PermuteDescriptor</a></div><div class="ttdoc">A PermuteDescriptor for the PermuteLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00102">Descriptors.hpp:102</a></div></div>
 204 <div class="ttc" id="namespacearmnn_html_a9c91b774c3089c55df77cc3a42da72de"><div class="ttname"><a href="namespacearmnn.html#a9c91b774c3089c55df77cc3a42da72de">armnn::CreateNetworkWithSoftmaxLayer</a></div><div class="ttdeci">INetworkPtr CreateNetworkWithSoftmaxLayer(const SoftmaxDescriptor &amp;descriptor, const TensorShape &amp;shape)</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l01357">QuantizerTest.cpp:1357</a></div></div>
 205 <div class="ttc" id="structarmnn_1_1_convolution2d_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_convolution2d_descriptor.html">armnn::Convolution2dDescriptor</a></div><div class="ttdoc">A Convolution2dDescriptor for the Convolution2dLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00392">Descriptors.hpp:392</a></div></div>
 206 <div class="ttc" id="classarmnn_1_1_network_html_ad55ff20f4c7e60c18b849e61f28f0e2e"><div class="ttname"><a href="classarmnn_1_1_network.html#ad55ff20f4c7e60c18b849e61f28f0e2e">armnn::Network::AddOutputLayer</a></div><div class="ttdeci">IConnectableLayer * AddOutputLayer(LayerBindingId id, const char *name=nullptr) override</div><div class="ttdef"><b>Definition:</b> <a href="_network_8cpp_source.html#l01222">Network.cpp:1222</a></div></div>
 207 <div class="ttc" id="namespacearmnn_html_a6fff4b4b1b5d4d37c9cf53d0e31c05dd"><div class="ttname"><a href="namespacearmnn.html#a6fff4b4b1b5d4d37c9cf53d0e31c05dd">armnn::CompleteLeakyReluNetwork</a></div><div class="ttdeci">void CompleteLeakyReluNetwork(INetwork *network, IConnectableLayer *activation, IConnectableLayer *layerUnderTest, const TensorInfo &amp;info)</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l01495">QuantizerTest.cpp:1495</a></div></div>
 208 <div class="ttc" id="classarmnn_1_1_i_output_slot_html_ac1835f8756a9f03c02fcf9664e3a0fce"><div class="ttname"><a href="classarmnn_1_1_i_output_slot.html#ac1835f8756a9f03c02fcf9664e3a0fce">armnn::IOutputSlot::Connect</a></div><div class="ttdeci">virtual int Connect(IInputSlot &amp;destination)=0</div></div>
 209 <div class="ttc" id="_file_only_profiling_decorator_tests_8cpp_html_a6560146509197f3e197d8d36f76c1347"><div class="ttname"><a href="_file_only_profiling_decorator_tests_8cpp.html#a6560146509197f3e197d8d36f76c1347">options</a></div><div class="ttdeci">armnn::Runtime::CreationOptions::ExternalProfilingOptions options</div><div class="ttdef"><b>Definition:</b> <a href="_file_only_profiling_decorator_tests_8cpp_source.html#l00045">FileOnlyProfilingDecoratorTests.cpp:45</a></div></div>
 210 <div class="ttc" id="namespacearmnn_html_a52cbff9d344ba4a1fe01d4da2c1f7ba2"><div class="ttname"><a href="namespacearmnn.html#a52cbff9d344ba4a1fe01d4da2c1f7ba2">armnn::SetupQuantize</a></div><div class="ttdeci">std::vector&lt; uint8_t &gt; SetupQuantize(float value)</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l02727">QuantizerTest.cpp:2727</a></div></div>
 211 <div class="ttc" id="structarmnn_1_1_space_to_batch_nd_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_space_to_batch_nd_descriptor.html">armnn::SpaceToBatchNdDescriptor</a></div><div class="ttdoc">A SpaceToBatchNdDescriptor for the SpaceToBatchNdLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00778">Descriptors.hpp:778</a></div></div>
 212 <div class="ttc" id="classarmnn_1_1_layer_visitor_base_html"><div class="ttname"><a href="classarmnn_1_1_layer_visitor_base.html">armnn::LayerVisitorBase</a></div><div class="ttdef"><b>Definition:</b> <a href="_layer_visitor_base_8hpp_source.html#l00025">LayerVisitorBase.hpp:25</a></div></div>
 213 <div class="ttc" id="namespacearmnn_html_afa7a0a639e2772ff2ced67d77be810c0"><div class="ttname"><a href="namespacearmnn.html#afa7a0a639e2772ff2ced67d77be810c0">armnn::TestQuantizeTransposeConvolution2d</a></div><div class="ttdeci">void TestQuantizeTransposeConvolution2d(bool useBiases)</div><div class="ttdef"><b>Definition:</b> <a href="_quantizer_test_8cpp_source.html#l02488">QuantizerTest.cpp:2488</a></div></div>
 214 <div class="ttc" id="classarmnn_1_1_tensor_info_html_a8b5d0f8a24e9d9238f412260a552acf8"><div class="ttname"><a href="classarmnn_1_1_tensor_info.html#a8b5d0f8a24e9d9238f412260a552acf8">armnn::TensorInfo::GetShape</a></div><div class="ttdeci">const TensorShape &amp; GetShape() const</div><div class="ttdef"><b>Definition:</b> <a href="_tensor_8hpp_source.html#l00088">Tensor.hpp:88</a></div></div>
 215 <div class="ttc" id="structarmnn_1_1_arg_min_max_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_arg_min_max_descriptor.html">armnn::ArgMinMaxDescriptor</a></div><div class="ttdoc">An ArgMinMaxDescriptor for ArgMinMaxLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00043">Descriptors.hpp:43</a></div></div>
 216 <div class="ttc" id="classarmnn_1_1_optional_base_html_a86b749ce2c4bc627fa8a1fcfaf0e314f"><div class="ttname"><a href="classarmnn_1_1_optional_base.html#a86b749ce2c4bc627fa8a1fcfaf0e314f">armnn::OptionalBase::has_value</a></div><div class="ttdeci">bool has_value() const noexcept</div><div class="ttdef"><b>Definition:</b> <a href="_optional_8hpp_source.html#l00053">Optional.hpp:53</a></div></div>
 217 <div class="ttc" id="namespacearmnn_html_ab8cf8f9fb6792e654c2d8d8382f6f01b"><div class="ttname"><a href="namespacearmnn.html#ab8cf8f9fb6792e654c2d8d8382f6f01b">armnn::LayerBindingId</a></div><div class="ttdeci">int LayerBindingId</div><div class="ttdoc">Type of identifiers for bindable layers (inputs, outputs). </div><div class="ttdef"><b>Definition:</b> <a href="_types_8hpp_source.html#l00168">Types.hpp:168</a></div></div>
 218 <div class="ttc" id="namespacearmnn_html_a81b5ff8545adad19a1c9d4ca076d552c"><div class="ttname"><a href="namespacearmnn.html#a81b5ff8545adad19a1c9d4ca076d552c">armnn::GetDataTypeName</a></div><div class="ttdeci">constexpr const char * GetDataTypeName(DataType dataType)</div><div class="ttdef"><b>Definition:</b> <a href="_types_utils_8hpp_source.html#l00165">TypesUtils.hpp:165</a></div></div>
 219 <div class="ttc" id="structarmnn_1_1_resize_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_resize_descriptor.html">armnn::ResizeDescriptor</a></div><div class="ttdoc">A ResizeDescriptor for the ResizeLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00724">Descriptors.hpp:724</a></div></div>
 220 <div class="ttc" id="namespacearmnn_html_a56297e0f7b215eea46c818cb7528d9eaa23b68da1de2b77d74da9da2635722a3e"><div class="ttname"><a href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaa23b68da1de2b77d74da9da2635722a3e">armnn::ActivationFunction::TanH</a></div></div>
 221 <div class="ttc" id="classarmnn_1_1profiling_1_1_profiling_guid_html"><div class="ttname"><a href="classarmnn_1_1profiling_1_1_profiling_guid.html">armnn::profiling::ProfilingGuid</a></div><div class="ttdef"><b>Definition:</b> <a href="_types_8hpp_source.html#l00249">Types.hpp:249</a></div></div>
 222 <div class="ttc" id="namespacearmnn_html_a56297e0f7b215eea46c818cb7528d9eaa32a843da6ea40ab3b17a3421ccdf671b"><div class="ttname"><a href="namespacearmnn.html#a56297e0f7b215eea46c818cb7528d9eaa32a843da6ea40ab3b17a3421ccdf671b">armnn::ActivationFunction::Linear</a></div></div>
 223 <div class="ttc" id="structarmnn_1_1_comparison_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_comparison_descriptor.html">armnn::ComparisonDescriptor</a></div><div class="ttdoc">A ComparisonDescriptor for the ComparisonLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00062">Descriptors.hpp:62</a></div></div>
 224 <div class="ttc" id="classarmnn_1_1_i_connectable_layer_html_a80ac4eda2e7f2757ec9dd96fc96dbd16"><div class="ttname"><a href="classarmnn_1_1_i_connectable_layer.html#a80ac4eda2e7f2757ec9dd96fc96dbd16">armnn::IConnectableLayer::GetOutputSlot</a></div><div class="ttdeci">virtual const IOutputSlot &amp; GetOutputSlot(unsigned int index) const =0</div></div>
 225 <div class="ttc" id="_i_network_8hpp_html"><div class="ttname"><a href="_i_network_8hpp.html">INetwork.hpp</a></div></div>
 226 <div class="ttc" id="classarmnn_1_1_tensor_info_html_a047ca888c43bd7fb5702853bf72410d0"><div class="ttname"><a href="classarmnn_1_1_tensor_info.html#a047ca888c43bd7fb5702853bf72410d0">armnn::TensorInfo::GetQuantizationScale</a></div><div class="ttdeci">float GetQuantizationScale() const</div><div class="ttdef"><b>Definition:</b> <a href="_tensor_8cpp_source.html#l00247">Tensor.cpp:247</a></div></div>
 227 <div class="ttc" id="structarmnn_1_1_quantizer_options_html"><div class="ttname"><a href="structarmnn_1_1_quantizer_options.html">armnn::QuantizerOptions</a></div><div class="ttdef"><b>Definition:</b> <a href="_i_network_quantizer_8hpp_source.html#l00015">INetworkQuantizer.hpp:15</a></div></div>
 228 <div class="ttc" id="structarmnn_1_1_origins_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_origins_descriptor.html">armnn::OriginsDescriptor</a></div><div class="ttdoc">An OriginsDescriptor for the ConcatLayer. Descriptor to configure the concatenation process...</div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l00147">Descriptors.hpp:147</a></div></div>
 229 <div class="ttc" id="classarmnn_1_1_optional_reference_switch_html_a77c7d528ac063d870b8c8426ec81c1c3"><div class="ttname"><a href="classarmnn_1_1_optional_reference_switch.html#a77c7d528ac063d870b8c8426ec81c1c3">armnn::OptionalReferenceSwitch&lt; std::is_reference&lt; T &gt;::value, T &gt;::value</a></div><div class="ttdeci">const T &amp; value() const</div><div class="ttdef"><b>Definition:</b> <a href="_optional_8hpp_source.html#l00146">Optional.hpp:146</a></div></div>
 230 <div class="ttc" id="namespacearmnn_html_ad1d5cce2d9e9a5d61c243e5c989112e0ad066db54b89b0912e7e7c6da51e2da51"><div class="ttname"><a href="namespacearmnn.html#ad1d5cce2d9e9a5d61c243e5c989112e0ad066db54b89b0912e7e7c6da51e2da51">armnn::DataLayout::NHWC</a></div></div>
 231 <div class="ttc" id="structarmnn_1_1_strided_slice_descriptor_html"><div class="ttname"><a href="structarmnn_1_1_strided_slice_descriptor.html">armnn::StridedSliceDescriptor</a></div><div class="ttdoc">A StridedSliceDescriptor for the StridedSliceLayer. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.html#l01002">Descriptors.hpp:1002</a></div></div>
 232 </div><!-- fragment --></div><!-- contents -->
 233 </div><!-- doc-content -->
 234 <!-- start footer part -->
 235 <div id="nav-path" class="navpath"><!-- id is needed for treeview function! -->
 236   <ul>
 237     <li class="navelem"><a class="el" href="dir_68267d1309a1af8e8297ef4c3efbcdba.html">src</a></li><li class="navelem"><a class="el" href="dir_e0a84d05c80a2ef4231141dcbbeac5c8.html">armnn</a></li><li class="navelem"><a class="el" href="dir_9d86fd1fbecbedf5bdb69c7e7235fe5f.html">test</a></li><li class="navelem"><a class="el" href="_quantizer_test_8cpp.html">QuantizerTest.cpp</a></li>
 238     <li class="footer">Generated on Fri Mar 13 2020 16:06:55 for ArmNN by
 239     <a href="http://www.doxygen.org/index.html">
 240     <img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.8.13 </li>
 241   </ul>
 242 </div>
 243 </body>
 244 </html>