| <a href="_l2_normalization_test_impl_8cpp.xhtml">Go to the documentation of this file.</a><div class="fragment"><div class="line"><a name="l00001"></a><span class="lineno"> 1</span> <span class="comment">//</span></div><div class="line"><a name="l00002"></a><span class="lineno"> 2</span> <span class="comment">// Copyright © 2017 Arm Ltd. All rights reserved.</span></div><div class="line"><a name="l00003"></a><span class="lineno"> 3</span> <span class="comment">// SPDX-License-Identifier: MIT</span></div><div class="line"><a name="l00004"></a><span class="lineno"> 4</span> <span class="comment">//</span></div><div class="line"><a name="l00005"></a><span class="lineno"> 5</span> </div><div class="line"><a name="l00006"></a><span class="lineno"> 6</span> <span class="preprocessor">#include "<a class="code" href="_l2_normalization_test_impl_8hpp.xhtml">L2NormalizationTestImpl.hpp</a>"</span></div><div class="line"><a name="l00007"></a><span class="lineno"> 7</span> </div><div class="line"><a name="l00008"></a><span class="lineno"> 8</span> <span class="preprocessor">#include <<a class="code" href="_quantize_helper_8hpp.xhtml">QuantizeHelper.hpp</a>></span></div><div class="line"><a name="l00009"></a><span class="lineno"> 9</span> <span class="preprocessor">#include <<a class="code" href="_resolve_type_8hpp.xhtml">ResolveType.hpp</a>></span></div><div class="line"><a name="l00010"></a><span class="lineno"> 10</span> </div><div class="line"><a name="l00011"></a><span class="lineno"> 11</span> <span class="preprocessor">#include <<a class="code" href="_tensor_utils_8hpp.xhtml">armnnUtils/TensorUtils.hpp</a>></span></div><div class="line"><a name="l00012"></a><span class="lineno"> 12</span> <span class="preprocessor">#include <<a class="code" href="_permute_8hpp.xhtml">armnnUtils/Permute.hpp</a>></span></div><div class="line"><a name="l00013"></a><span class="lineno"> 13</span> </div><div class="line"><a name="l00014"></a><span class="lineno"> 14</span> <span class="preprocessor">#include <<a class="code" href="_tensor_copy_utils_8hpp.xhtml">backendsCommon/test/TensorCopyUtils.hpp</a>></span></div><div class="line"><a name="l00015"></a><span class="lineno"> 15</span> <span class="preprocessor">#include <<a class="code" href="_workload_test_utils_8hpp.xhtml">backendsCommon/test/WorkloadTestUtils.hpp</a>></span></div><div class="line"><a name="l00016"></a><span class="lineno"> 16</span> </div><div class="line"><a name="l00017"></a><span class="lineno"> 17</span> <span class="preprocessor">#include <<a class="code" href="_tensor_helpers_8hpp.xhtml">test/TensorHelpers.hpp</a>></span></div><div class="line"><a name="l00018"></a><span class="lineno"> 18</span> </div><div class="line"><a name="l00019"></a><span class="lineno"> 19</span> <span class="keyword">namespace</span></div><div class="line"><a name="l00020"></a><span class="lineno"> 20</span> {</div><div class="line"><a name="l00021"></a><span class="lineno"> 21</span> </div><div class="line"><a name="l00022"></a><span class="lineno"> 22</span> <span class="keyword">template</span><armnn::DataType ArmnnType, <span class="keyword">typename</span> T = armnn::ResolveType<ArmnnType>></div><div class="line"><a name="l00023"></a><span class="lineno"> 23</span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<T, 4></a> L2NormalizationTestImpl(</div><div class="line"><a name="l00024"></a><span class="lineno"> 24</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00025"></a><span class="lineno"> 25</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00026"></a><span class="lineno"> 26</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.xhtml">armnn::TensorShape</a>& inputOutputTensorShape,</div><div class="line"><a name="l00027"></a><span class="lineno"> 27</span>  <span class="keywordtype">float</span> scale,</div><div class="line"><a name="l00028"></a><span class="lineno"> 28</span>  int32_t offset,</div><div class="line"><a name="l00029"></a><span class="lineno"> 29</span>  <span class="keyword">const</span> std::vector<float>& inputValues,</div><div class="line"><a name="l00030"></a><span class="lineno"> 30</span>  <span class="keywordtype">float</span> outScale,</div><div class="line"><a name="l00031"></a><span class="lineno"> 31</span>  int32_t outOffset,</div><div class="line"><a name="l00032"></a><span class="lineno"> 32</span>  <span class="keyword">const</span> std::vector<float>& expectedOutputValues,</div><div class="line"><a name="l00033"></a><span class="lineno"> 33</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout,</div><div class="line"><a name="l00034"></a><span class="lineno"> 34</span>  <span class="keywordtype">float</span> epsilon = 1e-12f)</div><div class="line"><a name="l00035"></a><span class="lineno"> 35</span> {</div><div class="line"><a name="l00036"></a><span class="lineno"> 36</span>  <a class="code" href="namespacearmnn.xhtml#a44affeeb090c3c6a3062830562672e84">IgnoreUnused</a>(memoryManager);</div><div class="line"><a name="l00037"></a><span class="lineno"> 37</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.xhtml">armnn::TensorInfo</a> inputTensorInfo(inputOutputTensorShape, ArmnnType, scale, offset);</div><div class="line"><a name="l00038"></a><span class="lineno"> 38</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.xhtml">armnn::TensorInfo</a> outputTensorInfo(inputOutputTensorShape, ArmnnType, outScale, outOffset);</div><div class="line"><a name="l00039"></a><span class="lineno"> 39</span> </div><div class="line"><a name="l00040"></a><span class="lineno"> 40</span>  <span class="comment">// at this point if we require it permute the input data</span></div><div class="line"><a name="l00041"></a><span class="lineno"> 41</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_permutation_vector.xhtml">armnn::PermutationVector</a> NCHWToNHWC = { 0, 3, 1, 2 };</div><div class="line"><a name="l00042"></a><span class="lineno"> 42</span>  std::vector<float> inputData = inputValues;</div><div class="line"><a name="l00043"></a><span class="lineno"> 43</span>  <span class="keywordflow">if</span> (layout == <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0ad066db54b89b0912e7e7c6da51e2da51">armnn::DataLayout::NHWC</a>)</div><div class="line"><a name="l00044"></a><span class="lineno"> 44</span>  {</div><div class="line"><a name="l00045"></a><span class="lineno"> 45</span>  std::vector<float> tmp(inputData.size());</div><div class="line"><a name="l00046"></a><span class="lineno"> 46</span>  <a class="code" href="namespacearmnn_utils.xhtml#af3c74017185773dd61d8ca6662d65d43">armnnUtils::Permute</a>(inputTensorInfo.GetShape(), NCHWToNHWC, inputData.data(), tmp.data(), <span class="keyword">sizeof</span>(float));</div><div class="line"><a name="l00047"></a><span class="lineno"> 47</span>  inputData = tmp;</div><div class="line"><a name="l00048"></a><span class="lineno"> 48</span>  }</div><div class="line"><a name="l00049"></a><span class="lineno"> 49</span> </div><div class="line"><a name="l00050"></a><span class="lineno"> 50</span>  <span class="keyword">auto</span> inputTensor = MakeTensor<T, 4>(inputTensorInfo,</div><div class="line"><a name="l00051"></a><span class="lineno"> 51</span>  armnnUtils::QuantizedVector<T>(inputData,</div><div class="line"><a name="l00052"></a><span class="lineno"> 52</span>  inputTensorInfo.GetQuantizationScale(),</div><div class="line"><a name="l00053"></a><span class="lineno"> 53</span>  inputTensorInfo.GetQuantizationOffset()));</div><div class="line"><a name="l00054"></a><span class="lineno"> 54</span> </div><div class="line"><a name="l00055"></a><span class="lineno"> 55</span>  std::vector<float> expectedOutputData = expectedOutputValues;</div><div class="line"><a name="l00056"></a><span class="lineno"> 56</span>  <span class="keywordflow">if</span> (layout == <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0ad066db54b89b0912e7e7c6da51e2da51">armnn::DataLayout::NHWC</a>)</div><div class="line"><a name="l00057"></a><span class="lineno"> 57</span>  {</div><div class="line"><a name="l00058"></a><span class="lineno"> 58</span>  std::vector<float> tmp(expectedOutputData.size());</div><div class="line"><a name="l00059"></a><span class="lineno"> 59</span>  <a class="code" href="namespacearmnn_utils.xhtml#af3c74017185773dd61d8ca6662d65d43">armnnUtils::Permute</a>(inputTensorInfo.GetShape(), NCHWToNHWC, expectedOutputData.data(), tmp.data(),</div><div class="line"><a name="l00060"></a><span class="lineno"> 60</span>  <span class="keyword">sizeof</span>(float));</div><div class="line"><a name="l00061"></a><span class="lineno"> 61</span>  expectedOutputData = tmp;</div><div class="line"><a name="l00062"></a><span class="lineno"> 62</span>  }</div><div class="line"><a name="l00063"></a><span class="lineno"> 63</span> </div><div class="line"><a name="l00064"></a><span class="lineno"> 64</span>  <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<T, 4></a> result(outputTensorInfo);</div><div class="line"><a name="l00065"></a><span class="lineno"> 65</span>  result.outputExpected =</div><div class="line"><a name="l00066"></a><span class="lineno"> 66</span>  MakeTensor<T, 4>(outputTensorInfo,</div><div class="line"><a name="l00067"></a><span class="lineno"> 67</span>  armnnUtils::QuantizedVector<T>(expectedOutputData,</div><div class="line"><a name="l00068"></a><span class="lineno"> 68</span>  outputTensorInfo.GetQuantizationScale(),</div><div class="line"><a name="l00069"></a><span class="lineno"> 69</span>  outputTensorInfo.GetQuantizationOffset()));</div><div class="line"><a name="l00070"></a><span class="lineno"> 70</span> </div><div class="line"><a name="l00071"></a><span class="lineno"> 71</span>  std::unique_ptr<armnn::ITensorHandle> inputHandle = workloadFactory.<a class="code" href="classarmnn_1_1_i_workload_factory.xhtml#a15c140be4ddceffee16436f009d3ed94">CreateTensorHandle</a>(inputTensorInfo);</div><div class="line"><a name="l00072"></a><span class="lineno"> 72</span>  std::unique_ptr<armnn::ITensorHandle> outputHandle = workloadFactory.<a class="code" href="classarmnn_1_1_i_workload_factory.xhtml#a15c140be4ddceffee16436f009d3ed94">CreateTensorHandle</a>(outputTensorInfo);</div><div class="line"><a name="l00073"></a><span class="lineno"> 73</span> </div><div class="line"><a name="l00074"></a><span class="lineno"> 74</span>  <a class="code" href="structarmnn_1_1_l2_normalization_queue_descriptor.xhtml">armnn::L2NormalizationQueueDescriptor</a> descriptor;</div><div class="line"><a name="l00075"></a><span class="lineno"> 75</span>  descriptor.<a class="code" href="structarmnn_1_1_queue_descriptor_with_parameters.xhtml#aad91b9bbf7aa365d304febe79a3d1333">m_Parameters</a>.<a class="code" href="structarmnn_1_1_l2_normalization_descriptor.xhtml#a11c821c7524251004a72ed13c510853c">m_Eps</a> = epsilon;</div><div class="line"><a name="l00076"></a><span class="lineno"> 76</span>  descriptor.<a class="code" href="structarmnn_1_1_queue_descriptor_with_parameters.xhtml#aad91b9bbf7aa365d304febe79a3d1333">m_Parameters</a>.<a class="code" href="structarmnn_1_1_l2_normalization_descriptor.xhtml#a6089e1ca91914015777ea780a513131a">m_DataLayout</a> = layout;</div><div class="line"><a name="l00077"></a><span class="lineno"> 77</span>  <a class="code" href="structarmnn_1_1_workload_info.xhtml">armnn::WorkloadInfo</a> <a class="code" href="namespacearmnn.xhtml#a4dc0adc6737b5944e7671bee71788407acaf9b6b99962bf5c2264824231d7a40c">info</a>;</div><div class="line"><a name="l00078"></a><span class="lineno"> 78</span> </div><div class="line"><a name="l00079"></a><span class="lineno"> 79</span>  AddInputToWorkload(descriptor, info, inputTensorInfo, inputHandle.get());</div><div class="line"><a name="l00080"></a><span class="lineno"> 80</span>  AddOutputToWorkload(descriptor, info, outputTensorInfo, outputHandle.get());</div><div class="line"><a name="l00081"></a><span class="lineno"> 81</span> </div><div class="line"><a name="l00082"></a><span class="lineno"> 82</span>  std::unique_ptr<armnn::IWorkload> workload = workloadFactory.<a class="code" href="classarmnn_1_1_i_workload_factory.xhtml#a3c86f886e36ce943f1ebc241a37f0413">CreateL2Normalization</a>(descriptor, info);</div><div class="line"><a name="l00083"></a><span class="lineno"> 83</span> </div><div class="line"><a name="l00084"></a><span class="lineno"> 84</span>  inputHandle->Allocate();</div><div class="line"><a name="l00085"></a><span class="lineno"> 85</span>  outputHandle->Allocate();</div><div class="line"><a name="l00086"></a><span class="lineno"> 86</span> </div><div class="line"><a name="l00087"></a><span class="lineno"> 87</span>  <a class="code" href="_tensor_copy_utils_8cpp.xhtml#ae15f1a3c55d2db87683577de9fa4437c">CopyDataToITensorHandle</a>(inputHandle.get(), &inputTensor[0][0][0][0]);</div><div class="line"><a name="l00088"></a><span class="lineno"> 88</span> </div><div class="line"><a name="l00089"></a><span class="lineno"> 89</span>  workload->PostAllocationConfigure();</div><div class="line"><a name="l00090"></a><span class="lineno"> 90</span>  ExecuteWorkload(*workload, memoryManager);</div><div class="line"><a name="l00091"></a><span class="lineno"> 91</span> </div><div class="line"><a name="l00092"></a><span class="lineno"> 92</span>  <a class="code" href="_tensor_copy_utils_8cpp.xhtml#a99b626c58a926dc7d6df78d22ec186c8">CopyDataFromITensorHandle</a>(&result.output[0][0][0][0], outputHandle.get());</div><div class="line"><a name="l00093"></a><span class="lineno"> 93</span> </div><div class="line"><a name="l00094"></a><span class="lineno"> 94</span>  <span class="keywordflow">return</span> result;</div><div class="line"><a name="l00095"></a><span class="lineno"> 95</span> }</div><div class="line"><a name="l00096"></a><span class="lineno"> 96</span> </div><div class="line"><a name="l00097"></a><span class="lineno"> 97</span> <span class="keywordtype">float</span> CalcInvL2Norm(std::initializer_list<float> elements)</div><div class="line"><a name="l00098"></a><span class="lineno"> 98</span> {</div><div class="line"><a name="l00099"></a><span class="lineno"> 99</span>  <span class="keyword">const</span> <span class="keywordtype">float</span> reduction = std::accumulate(elements.begin(), elements.end(), 0.0f,</div><div class="line"><a name="l00100"></a><span class="lineno"> 100</span>  [](<span class="keywordtype">float</span> acc, <span class="keywordtype">float</span> element) { <span class="keywordflow">return</span> acc + element * element; });</div><div class="line"><a name="l00101"></a><span class="lineno"> 101</span>  <span class="keywordflow">return</span> 1.0f / sqrtf(reduction);</div><div class="line"><a name="l00102"></a><span class="lineno"> 102</span> }</div><div class="line"><a name="l00103"></a><span class="lineno"> 103</span> </div><div class="line"><a name="l00104"></a><span class="lineno"> 104</span> <span class="keyword">template</span><armnn::DataType ArmnnType, <span class="keyword">typename</span> T = armnn::ResolveType<ArmnnType>></div><div class="line"><a name="l00105"></a><span class="lineno"> 105</span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<T, 4></a> L2NormalizationEpsilonTestCommon(</div><div class="line"><a name="l00106"></a><span class="lineno"> 106</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00107"></a><span class="lineno"> 107</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00108"></a><span class="lineno"> 108</span>  <span class="keywordtype">float</span> scale,</div><div class="line"><a name="l00109"></a><span class="lineno"> 109</span>  int32_t offset,</div><div class="line"><a name="l00110"></a><span class="lineno"> 110</span>  <span class="keywordtype">float</span> outScale,</div><div class="line"><a name="l00111"></a><span class="lineno"> 111</span>  int32_t outOffset,</div><div class="line"><a name="l00112"></a><span class="lineno"> 112</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout,</div><div class="line"><a name="l00113"></a><span class="lineno"> 113</span>  <span class="keywordtype">float</span> epsilon)</div><div class="line"><a name="l00114"></a><span class="lineno"> 114</span> {</div><div class="line"><a name="l00115"></a><span class="lineno"> 115</span>  <span class="comment">// Width: 1</span></div><div class="line"><a name="l00116"></a><span class="lineno"> 116</span>  <span class="comment">// Height: 1</span></div><div class="line"><a name="l00117"></a><span class="lineno"> 117</span>  <span class="comment">// Channels: 3</span></div><div class="line"><a name="l00118"></a><span class="lineno"> 118</span>  <span class="comment">// BatchSize: 1</span></div><div class="line"><a name="l00119"></a><span class="lineno"> 119</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> numberOfBatches = 1;</div><div class="line"><a name="l00120"></a><span class="lineno"> 120</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> numberOfChannels = 3;</div><div class="line"><a name="l00121"></a><span class="lineno"> 121</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> height = 1;</div><div class="line"><a name="l00122"></a><span class="lineno"> 122</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> width = 1;</div><div class="line"><a name="l00123"></a><span class="lineno"> 123</span> </div><div class="line"><a name="l00124"></a><span class="lineno"> 124</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.xhtml">armnn::TensorShape</a> inputOutputShape = <a class="code" href="namespacearmnn_utils.xhtml#ab53d94ea22b51c6bcdf9584644bd67bb">armnnUtils::GetTensorShape</a>(</div><div class="line"><a name="l00125"></a><span class="lineno"> 125</span>  numberOfBatches, numberOfChannels, height, width, layout);</div><div class="line"><a name="l00126"></a><span class="lineno"> 126</span> </div><div class="line"><a name="l00127"></a><span class="lineno"> 127</span>  <span class="comment">// 0.0000001^2 + 0.00000002^2 + 0.00000003^2 < 1e-12</span></div><div class="line"><a name="l00128"></a><span class="lineno"> 128</span>  std::vector<float> inputValues</div><div class="line"><a name="l00129"></a><span class="lineno"> 129</span>  {</div><div class="line"><a name="l00130"></a><span class="lineno"> 130</span>  <span class="comment">// Batch 0, Channel 0, Height (1) x Width (1)</span></div><div class="line"><a name="l00131"></a><span class="lineno"> 131</span>  0.00000001f,</div><div class="line"><a name="l00132"></a><span class="lineno"> 132</span> </div><div class="line"><a name="l00133"></a><span class="lineno"> 133</span>  <span class="comment">// Batch 0, Channel 1, Height (1) x Width (1)</span></div><div class="line"><a name="l00134"></a><span class="lineno"> 134</span>  0.00000002f,</div><div class="line"><a name="l00135"></a><span class="lineno"> 135</span> </div><div class="line"><a name="l00136"></a><span class="lineno"> 136</span>  <span class="comment">// Batch 0, Channel 2, Height (1) x Width (1)</span></div><div class="line"><a name="l00137"></a><span class="lineno"> 137</span>  0.00000003f,</div><div class="line"><a name="l00138"></a><span class="lineno"> 138</span>  };</div><div class="line"><a name="l00139"></a><span class="lineno"> 139</span> </div><div class="line"><a name="l00140"></a><span class="lineno"> 140</span>  <span class="keyword">const</span> <span class="keywordtype">float</span> approxInvL2Norm = 1.f / sqrtf(epsilon);</div><div class="line"><a name="l00141"></a><span class="lineno"> 141</span>  std::vector<float> expectedOutputValues</div><div class="line"><a name="l00142"></a><span class="lineno"> 142</span>  {</div><div class="line"><a name="l00143"></a><span class="lineno"> 143</span>  <span class="comment">// Batch 0, Channel 0, Height (1) x Width (1)</span></div><div class="line"><a name="l00144"></a><span class="lineno"> 144</span>  0.00000001f * approxInvL2Norm,</div><div class="line"><a name="l00145"></a><span class="lineno"> 145</span>  0.00000002f * approxInvL2Norm,</div><div class="line"><a name="l00146"></a><span class="lineno"> 146</span>  0.00000003f * approxInvL2Norm,</div><div class="line"><a name="l00147"></a><span class="lineno"> 147</span>  };</div><div class="line"><a name="l00148"></a><span class="lineno"> 148</span> </div><div class="line"><a name="l00149"></a><span class="lineno"> 149</span>  <span class="keywordflow">return</span> L2NormalizationTestImpl<ArmnnType>(</div><div class="line"><a name="l00150"></a><span class="lineno"> 150</span>  workloadFactory,</div><div class="line"><a name="l00151"></a><span class="lineno"> 151</span>  memoryManager,</div><div class="line"><a name="l00152"></a><span class="lineno"> 152</span>  inputOutputShape,</div><div class="line"><a name="l00153"></a><span class="lineno"> 153</span>  scale,</div><div class="line"><a name="l00154"></a><span class="lineno"> 154</span>  offset,</div><div class="line"><a name="l00155"></a><span class="lineno"> 155</span>  inputValues,</div><div class="line"><a name="l00156"></a><span class="lineno"> 156</span>  outScale,</div><div class="line"><a name="l00157"></a><span class="lineno"> 157</span>  outOffset,</div><div class="line"><a name="l00158"></a><span class="lineno"> 158</span>  expectedOutputValues,</div><div class="line"><a name="l00159"></a><span class="lineno"> 159</span>  layout,</div><div class="line"><a name="l00160"></a><span class="lineno"> 160</span>  epsilon);</div><div class="line"><a name="l00161"></a><span class="lineno"> 161</span> }</div><div class="line"><a name="l00162"></a><span class="lineno"> 162</span> </div><div class="line"><a name="l00163"></a><span class="lineno"> 163</span> </div><div class="line"><a name="l00164"></a><span class="lineno"> 164</span> <span class="keyword">template</span><armnn::DataType ArmnnType, <span class="keyword">typename</span> T = armnn::ResolveType<ArmnnType>></div><div class="line"><a name="l00165"></a><span class="lineno"> 165</span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<T, 4></a> L2Normalization1dTestCommon(</div><div class="line"><a name="l00166"></a><span class="lineno"> 166</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00167"></a><span class="lineno"> 167</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00168"></a><span class="lineno"> 168</span>  <span class="keywordtype">float</span> scale,</div><div class="line"><a name="l00169"></a><span class="lineno"> 169</span>  int32_t offset,</div><div class="line"><a name="l00170"></a><span class="lineno"> 170</span>  <span class="keywordtype">float</span> outScale,</div><div class="line"><a name="l00171"></a><span class="lineno"> 171</span>  int32_t outOffset,</div><div class="line"><a name="l00172"></a><span class="lineno"> 172</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00173"></a><span class="lineno"> 173</span> {</div><div class="line"><a name="l00174"></a><span class="lineno"> 174</span>  <span class="comment">// Width: 1</span></div><div class="line"><a name="l00175"></a><span class="lineno"> 175</span>  <span class="comment">// Height: 1</span></div><div class="line"><a name="l00176"></a><span class="lineno"> 176</span>  <span class="comment">// Channels: 10</span></div><div class="line"><a name="l00177"></a><span class="lineno"> 177</span>  <span class="comment">// BatchSize: 1</span></div><div class="line"><a name="l00178"></a><span class="lineno"> 178</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> numberOfBatches = 1;</div><div class="line"><a name="l00179"></a><span class="lineno"> 179</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> numberOfChannels = 10;</div><div class="line"><a name="l00180"></a><span class="lineno"> 180</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> height = 1;</div><div class="line"><a name="l00181"></a><span class="lineno"> 181</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> width = 1;</div><div class="line"><a name="l00182"></a><span class="lineno"> 182</span> </div><div class="line"><a name="l00183"></a><span class="lineno"> 183</span> </div><div class="line"><a name="l00184"></a><span class="lineno"> 184</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.xhtml">armnn::TensorShape</a> inputOutputShape = <a class="code" href="namespacearmnn_utils.xhtml#ab53d94ea22b51c6bcdf9584644bd67bb">armnnUtils::GetTensorShape</a>(</div><div class="line"><a name="l00185"></a><span class="lineno"> 185</span>  numberOfBatches, numberOfChannels, height, width, layout);</div><div class="line"><a name="l00186"></a><span class="lineno"> 186</span>  std::vector<float> inputValues</div><div class="line"><a name="l00187"></a><span class="lineno"> 187</span>  {</div><div class="line"><a name="l00188"></a><span class="lineno"> 188</span>  <span class="comment">// Batch 0, Channel 0, Height (1) x Width (1)</span></div><div class="line"><a name="l00189"></a><span class="lineno"> 189</span>  1.0f,</div><div class="line"><a name="l00190"></a><span class="lineno"> 190</span> </div><div class="line"><a name="l00191"></a><span class="lineno"> 191</span>  <span class="comment">// Batch 0, Channel 1, Height (1) x Width (1)</span></div><div class="line"><a name="l00192"></a><span class="lineno"> 192</span>  2.0f,</div><div class="line"><a name="l00193"></a><span class="lineno"> 193</span> </div><div class="line"><a name="l00194"></a><span class="lineno"> 194</span>  <span class="comment">// Batch 0, Channel 2, Height (1) x Width (1)</span></div><div class="line"><a name="l00195"></a><span class="lineno"> 195</span>  3.0f,</div><div class="line"><a name="l00196"></a><span class="lineno"> 196</span> </div><div class="line"><a name="l00197"></a><span class="lineno"> 197</span>  <span class="comment">// Batch 0, Channel 3, Height (1) x Width (1)</span></div><div class="line"><a name="l00198"></a><span class="lineno"> 198</span>  4.0f,</div><div class="line"><a name="l00199"></a><span class="lineno"> 199</span> </div><div class="line"><a name="l00200"></a><span class="lineno"> 200</span>  <span class="comment">// Batch 0, Channel 4, Height (1) x Width (1)</span></div><div class="line"><a name="l00201"></a><span class="lineno"> 201</span>  5.0f,</div><div class="line"><a name="l00202"></a><span class="lineno"> 202</span> </div><div class="line"><a name="l00203"></a><span class="lineno"> 203</span>  <span class="comment">// Batch 0, Channel 5, Height (1) x Width (1)</span></div><div class="line"><a name="l00204"></a><span class="lineno"> 204</span>  6.0f,</div><div class="line"><a name="l00205"></a><span class="lineno"> 205</span> </div><div class="line"><a name="l00206"></a><span class="lineno"> 206</span>  <span class="comment">// Batch 0, Channel 6, Height (1) x Width (1)</span></div><div class="line"><a name="l00207"></a><span class="lineno"> 207</span>  7.0f,</div><div class="line"><a name="l00208"></a><span class="lineno"> 208</span> </div><div class="line"><a name="l00209"></a><span class="lineno"> 209</span>  <span class="comment">// Batch 0, Channel 7, Height (1) x Width (1)</span></div><div class="line"><a name="l00210"></a><span class="lineno"> 210</span>  8.0f,</div><div class="line"><a name="l00211"></a><span class="lineno"> 211</span> </div><div class="line"><a name="l00212"></a><span class="lineno"> 212</span>  <span class="comment">// Batch 0, Channel 8, Height (1) x Width (1)</span></div><div class="line"><a name="l00213"></a><span class="lineno"> 213</span>  9.0f,</div><div class="line"><a name="l00214"></a><span class="lineno"> 214</span> </div><div class="line"><a name="l00215"></a><span class="lineno"> 215</span>  <span class="comment">// Batch 0, Channel 9, Height (1) x Width (1)</span></div><div class="line"><a name="l00216"></a><span class="lineno"> 216</span>  10.0f</div><div class="line"><a name="l00217"></a><span class="lineno"> 217</span>  };</div><div class="line"><a name="l00218"></a><span class="lineno"> 218</span>  <span class="keyword">const</span> <span class="keywordtype">float</span> approxInvL2Norm = 0.050964719f;</div><div class="line"><a name="l00219"></a><span class="lineno"> 219</span>  std::vector<float> expectedOutputValues</div><div class="line"><a name="l00220"></a><span class="lineno"> 220</span>  {</div><div class="line"><a name="l00221"></a><span class="lineno"> 221</span>  <span class="comment">// Batch 0, Channel 0, Height (1) x Width (1)</span></div><div class="line"><a name="l00222"></a><span class="lineno"> 222</span>  1.0f * approxInvL2Norm,</div><div class="line"><a name="l00223"></a><span class="lineno"> 223</span>  2.0f * approxInvL2Norm,</div><div class="line"><a name="l00224"></a><span class="lineno"> 224</span>  3.0f * approxInvL2Norm,</div><div class="line"><a name="l00225"></a><span class="lineno"> 225</span>  4.0f * approxInvL2Norm,</div><div class="line"><a name="l00226"></a><span class="lineno"> 226</span>  5.0f * approxInvL2Norm,</div><div class="line"><a name="l00227"></a><span class="lineno"> 227</span>  6.0f * approxInvL2Norm,</div><div class="line"><a name="l00228"></a><span class="lineno"> 228</span>  7.0f * approxInvL2Norm,</div><div class="line"><a name="l00229"></a><span class="lineno"> 229</span>  8.0f * approxInvL2Norm,</div><div class="line"><a name="l00230"></a><span class="lineno"> 230</span>  9.0f * approxInvL2Norm,</div><div class="line"><a name="l00231"></a><span class="lineno"> 231</span>  10.0f * approxInvL2Norm</div><div class="line"><a name="l00232"></a><span class="lineno"> 232</span>  };</div><div class="line"><a name="l00233"></a><span class="lineno"> 233</span> </div><div class="line"><a name="l00234"></a><span class="lineno"> 234</span> </div><div class="line"><a name="l00235"></a><span class="lineno"> 235</span>  <span class="keywordflow">return</span> L2NormalizationTestImpl<ArmnnType>(</div><div class="line"><a name="l00236"></a><span class="lineno"> 236</span>  workloadFactory,</div><div class="line"><a name="l00237"></a><span class="lineno"> 237</span>  memoryManager,</div><div class="line"><a name="l00238"></a><span class="lineno"> 238</span>  inputOutputShape,</div><div class="line"><a name="l00239"></a><span class="lineno"> 239</span>  scale,</div><div class="line"><a name="l00240"></a><span class="lineno"> 240</span>  offset,</div><div class="line"><a name="l00241"></a><span class="lineno"> 241</span>  inputValues,</div><div class="line"><a name="l00242"></a><span class="lineno"> 242</span>  outScale,</div><div class="line"><a name="l00243"></a><span class="lineno"> 243</span>  outOffset,</div><div class="line"><a name="l00244"></a><span class="lineno"> 244</span>  expectedOutputValues,</div><div class="line"><a name="l00245"></a><span class="lineno"> 245</span>  layout);</div><div class="line"><a name="l00246"></a><span class="lineno"> 246</span> }</div><div class="line"><a name="l00247"></a><span class="lineno"> 247</span> </div><div class="line"><a name="l00248"></a><span class="lineno"> 248</span> <span class="keyword">template</span><armnn::DataType ArmnnType, <span class="keyword">typename</span> T = armnn::ResolveType<ArmnnType>></div><div class="line"><a name="l00249"></a><span class="lineno"> 249</span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<T, 4></a> L2Normalization2dTestCommon(</div><div class="line"><a name="l00250"></a><span class="lineno"> 250</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00251"></a><span class="lineno"> 251</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00252"></a><span class="lineno"> 252</span>  <span class="keywordtype">float</span> scale,</div><div class="line"><a name="l00253"></a><span class="lineno"> 253</span>  int32_t offset,</div><div class="line"><a name="l00254"></a><span class="lineno"> 254</span>  <span class="keywordtype">float</span> outScale,</div><div class="line"><a name="l00255"></a><span class="lineno"> 255</span>  int32_t outOffset,</div><div class="line"><a name="l00256"></a><span class="lineno"> 256</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00257"></a><span class="lineno"> 257</span> {</div><div class="line"><a name="l00258"></a><span class="lineno"> 258</span>  <span class="comment">// Width: 5</span></div><div class="line"><a name="l00259"></a><span class="lineno"> 259</span>  <span class="comment">// Height: 1</span></div><div class="line"><a name="l00260"></a><span class="lineno"> 260</span>  <span class="comment">// Channels: 2</span></div><div class="line"><a name="l00261"></a><span class="lineno"> 261</span>  <span class="comment">// BatchSize: 1</span></div><div class="line"><a name="l00262"></a><span class="lineno"> 262</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> numberOfBatches = 1;</div><div class="line"><a name="l00263"></a><span class="lineno"> 263</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> numberOfChannels = 2;</div><div class="line"><a name="l00264"></a><span class="lineno"> 264</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> height = 1;</div><div class="line"><a name="l00265"></a><span class="lineno"> 265</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> width = 5;</div><div class="line"><a name="l00266"></a><span class="lineno"> 266</span> </div><div class="line"><a name="l00267"></a><span class="lineno"> 267</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.xhtml">armnn::TensorShape</a> inputOutputShape = <a class="code" href="namespacearmnn_utils.xhtml#ab53d94ea22b51c6bcdf9584644bd67bb">armnnUtils::GetTensorShape</a>(</div><div class="line"><a name="l00268"></a><span class="lineno"> 268</span>  numberOfBatches, numberOfChannels, height, width, layout);</div><div class="line"><a name="l00269"></a><span class="lineno"> 269</span>  std::vector<float> inputValues</div><div class="line"><a name="l00270"></a><span class="lineno"> 270</span>  {</div><div class="line"><a name="l00271"></a><span class="lineno"> 271</span>  <span class="comment">// Batch 0, Channel 0, Height (1) x Width (5)</span></div><div class="line"><a name="l00272"></a><span class="lineno"> 272</span>  1.0f, 3.0f, 5.0f, 7.0f, 9.0f,</div><div class="line"><a name="l00273"></a><span class="lineno"> 273</span> </div><div class="line"><a name="l00274"></a><span class="lineno"> 274</span>  <span class="comment">// Batch 0, Channel 1, Height (1) x Width (5)</span></div><div class="line"><a name="l00275"></a><span class="lineno"> 275</span>  2.0f, 4.0f, 6.0f, 8.0f, 10.0f</div><div class="line"><a name="l00276"></a><span class="lineno"> 276</span>  };</div><div class="line"><a name="l00277"></a><span class="lineno"> 277</span>  std::vector<float> expectedOutputValues</div><div class="line"><a name="l00278"></a><span class="lineno"> 278</span>  {</div><div class="line"><a name="l00279"></a><span class="lineno"> 279</span>  <span class="comment">// Batch 0, Channel 0, Height (1) x Width (5)</span></div><div class="line"><a name="l00280"></a><span class="lineno"> 280</span>  1.0f * CalcInvL2Norm({ 1.0f, 2.0f }),</div><div class="line"><a name="l00281"></a><span class="lineno"> 281</span>  3.0f * CalcInvL2Norm({ 3.0f, 4.0f }),</div><div class="line"><a name="l00282"></a><span class="lineno"> 282</span>  5.0f * CalcInvL2Norm({ 5.0f, 6.0f }),</div><div class="line"><a name="l00283"></a><span class="lineno"> 283</span>  7.0f * CalcInvL2Norm({ 7.0f, 8.0f }),</div><div class="line"><a name="l00284"></a><span class="lineno"> 284</span>  9.0f * CalcInvL2Norm({ 9.0f, 10.0f }),</div><div class="line"><a name="l00285"></a><span class="lineno"> 285</span> </div><div class="line"><a name="l00286"></a><span class="lineno"> 286</span>  <span class="comment">// Batch 0, Channel 1, Height (1) x Width (5)</span></div><div class="line"><a name="l00287"></a><span class="lineno"> 287</span>  2.0f * CalcInvL2Norm({ 1.0f, 2.0f }),</div><div class="line"><a name="l00288"></a><span class="lineno"> 288</span>  4.0f * CalcInvL2Norm({ 3.0f, 4.0f }),</div><div class="line"><a name="l00289"></a><span class="lineno"> 289</span>  6.0f * CalcInvL2Norm({ 5.0f, 6.0f }),</div><div class="line"><a name="l00290"></a><span class="lineno"> 290</span>  8.0f * CalcInvL2Norm({ 7.0f, 8.0f }),</div><div class="line"><a name="l00291"></a><span class="lineno"> 291</span>  10.0f * CalcInvL2Norm({ 9.0f, 10.0f })</div><div class="line"><a name="l00292"></a><span class="lineno"> 292</span>  };</div><div class="line"><a name="l00293"></a><span class="lineno"> 293</span> </div><div class="line"><a name="l00294"></a><span class="lineno"> 294</span>  <span class="keywordflow">return</span> L2NormalizationTestImpl<ArmnnType>(</div><div class="line"><a name="l00295"></a><span class="lineno"> 295</span>  workloadFactory,</div><div class="line"><a name="l00296"></a><span class="lineno"> 296</span>  memoryManager,</div><div class="line"><a name="l00297"></a><span class="lineno"> 297</span>  inputOutputShape,</div><div class="line"><a name="l00298"></a><span class="lineno"> 298</span>  scale,</div><div class="line"><a name="l00299"></a><span class="lineno"> 299</span>  offset,</div><div class="line"><a name="l00300"></a><span class="lineno"> 300</span>  inputValues,</div><div class="line"><a name="l00301"></a><span class="lineno"> 301</span>  outScale,</div><div class="line"><a name="l00302"></a><span class="lineno"> 302</span>  outOffset,</div><div class="line"><a name="l00303"></a><span class="lineno"> 303</span>  expectedOutputValues,</div><div class="line"><a name="l00304"></a><span class="lineno"> 304</span>  layout);</div><div class="line"><a name="l00305"></a><span class="lineno"> 305</span> }</div><div class="line"><a name="l00306"></a><span class="lineno"> 306</span> </div><div class="line"><a name="l00307"></a><span class="lineno"> 307</span> <span class="keyword">template</span><armnn::DataType ArmnnType, <span class="keyword">typename</span> T = armnn::ResolveType<ArmnnType>></div><div class="line"><a name="l00308"></a><span class="lineno"> 308</span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<T, 4></a> L2Normalization3dTestCommon(</div><div class="line"><a name="l00309"></a><span class="lineno"> 309</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00310"></a><span class="lineno"> 310</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00311"></a><span class="lineno"> 311</span>  <span class="keywordtype">float</span> scale,</div><div class="line"><a name="l00312"></a><span class="lineno"> 312</span>  int32_t offset,</div><div class="line"><a name="l00313"></a><span class="lineno"> 313</span>  <span class="keywordtype">float</span> outScale,</div><div class="line"><a name="l00314"></a><span class="lineno"> 314</span>  int32_t outOffset,</div><div class="line"><a name="l00315"></a><span class="lineno"> 315</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00316"></a><span class="lineno"> 316</span> {</div><div class="line"><a name="l00317"></a><span class="lineno"> 317</span>  <span class="comment">// Width: 3</span></div><div class="line"><a name="l00318"></a><span class="lineno"> 318</span>  <span class="comment">// Height: 4</span></div><div class="line"><a name="l00319"></a><span class="lineno"> 319</span>  <span class="comment">// Channels: 2</span></div><div class="line"><a name="l00320"></a><span class="lineno"> 320</span>  <span class="comment">// BatchSize: 1</span></div><div class="line"><a name="l00321"></a><span class="lineno"> 321</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> numberOfBatches = 1;</div><div class="line"><a name="l00322"></a><span class="lineno"> 322</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> numberOfChannels = 2;</div><div class="line"><a name="l00323"></a><span class="lineno"> 323</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> height = 4;</div><div class="line"><a name="l00324"></a><span class="lineno"> 324</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> width = 3;</div><div class="line"><a name="l00325"></a><span class="lineno"> 325</span> </div><div class="line"><a name="l00326"></a><span class="lineno"> 326</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.xhtml">armnn::TensorShape</a> inputOutputShape = <a class="code" href="namespacearmnn_utils.xhtml#ab53d94ea22b51c6bcdf9584644bd67bb">armnnUtils::GetTensorShape</a>(</div><div class="line"><a name="l00327"></a><span class="lineno"> 327</span>  numberOfBatches, numberOfChannels, height, width, layout);</div><div class="line"><a name="l00328"></a><span class="lineno"> 328</span>  std::vector<float> inputValues</div><div class="line"><a name="l00329"></a><span class="lineno"> 329</span>  {</div><div class="line"><a name="l00330"></a><span class="lineno"> 330</span>  <span class="comment">// Batch 0, Channel 0, Height (4) x Width (3)</span></div><div class="line"><a name="l00331"></a><span class="lineno"> 331</span>  119.0f, 21.0f, 150.0f,</div><div class="line"><a name="l00332"></a><span class="lineno"> 332</span>  149.0f, 32.0f, 179.0f,</div><div class="line"><a name="l00333"></a><span class="lineno"> 333</span>  15.0f, 227.0f, 141.0f,</div><div class="line"><a name="l00334"></a><span class="lineno"> 334</span>  147.0f, 199.0f, 220.0f,</div><div class="line"><a name="l00335"></a><span class="lineno"> 335</span> </div><div class="line"><a name="l00336"></a><span class="lineno"> 336</span>  <span class="comment">// Batch 0, Channel 1, Height (4) x Width (3)</span></div><div class="line"><a name="l00337"></a><span class="lineno"> 337</span>  110.0f, 140.0f, 73.0f,</div><div class="line"><a name="l00338"></a><span class="lineno"> 338</span>  211.0f, 212.0f, 89.0f,</div><div class="line"><a name="l00339"></a><span class="lineno"> 339</span>  24.0f, 138.0f, 188.0f,</div><div class="line"><a name="l00340"></a><span class="lineno"> 340</span>  162.0f, 12.0f, 161.0f</div><div class="line"><a name="l00341"></a><span class="lineno"> 341</span>  };</div><div class="line"><a name="l00342"></a><span class="lineno"> 342</span>  std::vector<float> expectedOutputValues</div><div class="line"><a name="l00343"></a><span class="lineno"> 343</span>  {</div><div class="line"><a name="l00344"></a><span class="lineno"> 344</span>  <span class="comment">// Batch 0, Channel 0, Height (4) x Width (3)</span></div><div class="line"><a name="l00345"></a><span class="lineno"> 345</span>  119.0f * CalcInvL2Norm({ 119.0f, 110.0f }),</div><div class="line"><a name="l00346"></a><span class="lineno"> 346</span>  21.0f * CalcInvL2Norm({ 21.0f, 140.0f }),</div><div class="line"><a name="l00347"></a><span class="lineno"> 347</span>  150.0f * CalcInvL2Norm({ 150.0f, 73.0f }),</div><div class="line"><a name="l00348"></a><span class="lineno"> 348</span>  149.0f * CalcInvL2Norm({ 149.0f, 211.0f }),</div><div class="line"><a name="l00349"></a><span class="lineno"> 349</span>  32.0f * CalcInvL2Norm({ 32.0f, 212.0f }),</div><div class="line"><a name="l00350"></a><span class="lineno"> 350</span>  179.0f * CalcInvL2Norm({ 179.0f, 89.0f }),</div><div class="line"><a name="l00351"></a><span class="lineno"> 351</span>  15.0f * CalcInvL2Norm({ 15.0f, 24.0f }),</div><div class="line"><a name="l00352"></a><span class="lineno"> 352</span>  227.0f * CalcInvL2Norm({ 227.0f, 138.0f }),</div><div class="line"><a name="l00353"></a><span class="lineno"> 353</span>  141.0f * CalcInvL2Norm({ 141.0f, 188.0f }),</div><div class="line"><a name="l00354"></a><span class="lineno"> 354</span>  147.0f * CalcInvL2Norm({ 147.0f, 162.0f }),</div><div class="line"><a name="l00355"></a><span class="lineno"> 355</span>  199.0f * CalcInvL2Norm({ 199.0f, 12.0f }),</div><div class="line"><a name="l00356"></a><span class="lineno"> 356</span>  220.0f * CalcInvL2Norm({ 220.0f, 161.0f }),</div><div class="line"><a name="l00357"></a><span class="lineno"> 357</span> </div><div class="line"><a name="l00358"></a><span class="lineno"> 358</span>  <span class="comment">// Batch 0, Channel 1, Height (4) x Width (3)</span></div><div class="line"><a name="l00359"></a><span class="lineno"> 359</span>  110.0f * CalcInvL2Norm({ 119.0f, 110.0f }),</div><div class="line"><a name="l00360"></a><span class="lineno"> 360</span>  140.0f * CalcInvL2Norm({ 21.0f, 140.0f }),</div><div class="line"><a name="l00361"></a><span class="lineno"> 361</span>  73.0f * CalcInvL2Norm({ 150.0f, 73.0f }),</div><div class="line"><a name="l00362"></a><span class="lineno"> 362</span>  211.0f * CalcInvL2Norm({ 149.0f, 211.0f }),</div><div class="line"><a name="l00363"></a><span class="lineno"> 363</span>  212.0f * CalcInvL2Norm({ 32.0f, 212.0f }),</div><div class="line"><a name="l00364"></a><span class="lineno"> 364</span>  89.0f * CalcInvL2Norm({ 179.0f, 89.0f }),</div><div class="line"><a name="l00365"></a><span class="lineno"> 365</span>  24.0f * CalcInvL2Norm({ 15.0f, 24.0f }),</div><div class="line"><a name="l00366"></a><span class="lineno"> 366</span>  138.0f * CalcInvL2Norm({ 227.0f, 138.0f }),</div><div class="line"><a name="l00367"></a><span class="lineno"> 367</span>  188.0f * CalcInvL2Norm({ 141.0f, 188.0f }),</div><div class="line"><a name="l00368"></a><span class="lineno"> 368</span>  162.0f * CalcInvL2Norm({ 147.0f, 162.0f }),</div><div class="line"><a name="l00369"></a><span class="lineno"> 369</span>  12.0f * CalcInvL2Norm({ 199.0f, 12.0f }),</div><div class="line"><a name="l00370"></a><span class="lineno"> 370</span>  161.0f * CalcInvL2Norm({ 220.0f, 161.0f })</div><div class="line"><a name="l00371"></a><span class="lineno"> 371</span>  };</div><div class="line"><a name="l00372"></a><span class="lineno"> 372</span> </div><div class="line"><a name="l00373"></a><span class="lineno"> 373</span>  <span class="keywordflow">return</span> L2NormalizationTestImpl<ArmnnType>(</div><div class="line"><a name="l00374"></a><span class="lineno"> 374</span>  workloadFactory,</div><div class="line"><a name="l00375"></a><span class="lineno"> 375</span>  memoryManager,</div><div class="line"><a name="l00376"></a><span class="lineno"> 376</span>  inputOutputShape,</div><div class="line"><a name="l00377"></a><span class="lineno"> 377</span>  scale,</div><div class="line"><a name="l00378"></a><span class="lineno"> 378</span>  offset,</div><div class="line"><a name="l00379"></a><span class="lineno"> 379</span>  inputValues,</div><div class="line"><a name="l00380"></a><span class="lineno"> 380</span>  outScale,</div><div class="line"><a name="l00381"></a><span class="lineno"> 381</span>  outOffset,</div><div class="line"><a name="l00382"></a><span class="lineno"> 382</span>  expectedOutputValues,</div><div class="line"><a name="l00383"></a><span class="lineno"> 383</span>  layout);</div><div class="line"><a name="l00384"></a><span class="lineno"> 384</span> }</div><div class="line"><a name="l00385"></a><span class="lineno"> 385</span> </div><div class="line"><a name="l00386"></a><span class="lineno"> 386</span> <span class="keyword">template</span><armnn::DataType ArmnnType, <span class="keyword">typename</span> T = armnn::ResolveType<ArmnnType>></div><div class="line"><a name="l00387"></a><span class="lineno"> 387</span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<T, 4></a> L2Normalization4dTestCommon(</div><div class="line"><a name="l00388"></a><span class="lineno"> 388</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00389"></a><span class="lineno"> 389</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00390"></a><span class="lineno"> 390</span>  <span class="keywordtype">float</span> scale,</div><div class="line"><a name="l00391"></a><span class="lineno"> 391</span>  int32_t offset,</div><div class="line"><a name="l00392"></a><span class="lineno"> 392</span>  <span class="keywordtype">float</span> outScale,</div><div class="line"><a name="l00393"></a><span class="lineno"> 393</span>  int32_t outOffset,</div><div class="line"><a name="l00394"></a><span class="lineno"> 394</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00395"></a><span class="lineno"> 395</span> {</div><div class="line"><a name="l00396"></a><span class="lineno"> 396</span>  <span class="comment">// Width: 3</span></div><div class="line"><a name="l00397"></a><span class="lineno"> 397</span>  <span class="comment">// Height: 4</span></div><div class="line"><a name="l00398"></a><span class="lineno"> 398</span>  <span class="comment">// Channels: 3</span></div><div class="line"><a name="l00399"></a><span class="lineno"> 399</span>  <span class="comment">// BatchSize: 2</span></div><div class="line"><a name="l00400"></a><span class="lineno"> 400</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> numberOfBatches = 2;</div><div class="line"><a name="l00401"></a><span class="lineno"> 401</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> numberOfChannels = 3;</div><div class="line"><a name="l00402"></a><span class="lineno"> 402</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> height = 4;</div><div class="line"><a name="l00403"></a><span class="lineno"> 403</span>  <span class="keywordtype">unsigned</span> <span class="keywordtype">int</span> width = 3;</div><div class="line"><a name="l00404"></a><span class="lineno"> 404</span> </div><div class="line"><a name="l00405"></a><span class="lineno"> 405</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.xhtml">armnn::TensorShape</a> inputOutputShape = <a class="code" href="namespacearmnn_utils.xhtml#ab53d94ea22b51c6bcdf9584644bd67bb">armnnUtils::GetTensorShape</a>(</div><div class="line"><a name="l00406"></a><span class="lineno"> 406</span>  numberOfBatches, numberOfChannels, height, width, layout);</div><div class="line"><a name="l00407"></a><span class="lineno"> 407</span>  std::vector<float> inputValues</div><div class="line"><a name="l00408"></a><span class="lineno"> 408</span>  {</div><div class="line"><a name="l00409"></a><span class="lineno"> 409</span>  <span class="comment">// Batch 0, Channel 0, Height (4) x Width (3)</span></div><div class="line"><a name="l00410"></a><span class="lineno"> 410</span>  235.0f, 46.0f, 178.0f,</div><div class="line"><a name="l00411"></a><span class="lineno"> 411</span>  100.0f, 123.0f, 19.0f,</div><div class="line"><a name="l00412"></a><span class="lineno"> 412</span>  172.0f, 74.0f, 250.0f,</div><div class="line"><a name="l00413"></a><span class="lineno"> 413</span>  6.0f, 195.0f, 80.0f,</div><div class="line"><a name="l00414"></a><span class="lineno"> 414</span> </div><div class="line"><a name="l00415"></a><span class="lineno"> 415</span>  <span class="comment">// Batch 0, Channel 1, Height (4) x Width (3)</span></div><div class="line"><a name="l00416"></a><span class="lineno"> 416</span>  113.0f, 95.0f, 202.0f,</div><div class="line"><a name="l00417"></a><span class="lineno"> 417</span>  77.0f, 114.0f, 71.0f,</div><div class="line"><a name="l00418"></a><span class="lineno"> 418</span>  122.0f, 246.0f, 166.0f,</div><div class="line"><a name="l00419"></a><span class="lineno"> 419</span>  82.0f, 28.0f, 37.0f,</div><div class="line"><a name="l00420"></a><span class="lineno"> 420</span> </div><div class="line"><a name="l00421"></a><span class="lineno"> 421</span>  <span class="comment">// Batch 0, Channel 2, Height (4) x Width (3)</span></div><div class="line"><a name="l00422"></a><span class="lineno"> 422</span>  56.0f, 170.0f, 162.0f,</div><div class="line"><a name="l00423"></a><span class="lineno"> 423</span>  194.0f, 89.0f, 254.0f,</div><div class="line"><a name="l00424"></a><span class="lineno"> 424</span>  12.0f, 209.0f, 200.0f,</div><div class="line"><a name="l00425"></a><span class="lineno"> 425</span>  1.0f, 64.0f, 54.0f,</div><div class="line"><a name="l00426"></a><span class="lineno"> 426</span> </div><div class="line"><a name="l00427"></a><span class="lineno"> 427</span>  <span class="comment">// Batch 1, Channel 0, Height (4) x Width (3)</span></div><div class="line"><a name="l00428"></a><span class="lineno"> 428</span>  67.0f, 90.0f, 49.0f,</div><div class="line"><a name="l00429"></a><span class="lineno"> 429</span>  7.0f, 163.0f, 18.0f,</div><div class="line"><a name="l00430"></a><span class="lineno"> 430</span>  25.0f, 117.0f, 103.0f,</div><div class="line"><a name="l00431"></a><span class="lineno"> 431</span>  247.0f, 59.0f, 189.0f,</div><div class="line"><a name="l00432"></a><span class="lineno"> 432</span> </div><div class="line"><a name="l00433"></a><span class="lineno"> 433</span>  <span class="comment">// Batch 1, Channel 1, Height (4) x Width (3)</span></div><div class="line"><a name="l00434"></a><span class="lineno"> 434</span>  239.0f, 104.0f, 199.0f,</div><div class="line"><a name="l00435"></a><span class="lineno"> 435</span>  17.0f, 124.0f, 153.0f,</div><div class="line"><a name="l00436"></a><span class="lineno"> 436</span>  222.0f, 217.0f, 75.0f,</div><div class="line"><a name="l00437"></a><span class="lineno"> 437</span>  32.0f, 126.0f, 21.0f,</div><div class="line"><a name="l00438"></a><span class="lineno"> 438</span> </div><div class="line"><a name="l00439"></a><span class="lineno"> 439</span>  <span class="comment">// Batch 1, Channel 2, Height (4) x Width (3)</span></div><div class="line"><a name="l00440"></a><span class="lineno"> 440</span>  97.0f, 145.0f, 215.0f,</div><div class="line"><a name="l00441"></a><span class="lineno"> 441</span>  115.0f, 116.0f, 238.0f,</div><div class="line"><a name="l00442"></a><span class="lineno"> 442</span>  226.0f, 16.0f, 132.0f,</div><div class="line"><a name="l00443"></a><span class="lineno"> 443</span>  92.0f, 125.0f, 88.0f</div><div class="line"><a name="l00444"></a><span class="lineno"> 444</span>  };</div><div class="line"><a name="l00445"></a><span class="lineno"> 445</span>  std::vector<float> expectedOutputValues</div><div class="line"><a name="l00446"></a><span class="lineno"> 446</span>  {</div><div class="line"><a name="l00447"></a><span class="lineno"> 447</span>  <span class="comment">// Batch 0, Channel 0, Height (4) x Width (3)</span></div><div class="line"><a name="l00448"></a><span class="lineno"> 448</span>  235.0f * CalcInvL2Norm({ 235.0f, 113.0f, 56.0f }),</div><div class="line"><a name="l00449"></a><span class="lineno"> 449</span>  46.0f * CalcInvL2Norm({ 46.0f, 95.0f, 170.0f }),</div><div class="line"><a name="l00450"></a><span class="lineno"> 450</span>  178.0f * CalcInvL2Norm({ 178.0f, 202.0F, 162.0f }),</div><div class="line"><a name="l00451"></a><span class="lineno"> 451</span>  100.0f * CalcInvL2Norm({ 100.0f, 77.0f, 194.0f }),</div><div class="line"><a name="l00452"></a><span class="lineno"> 452</span>  123.0f * CalcInvL2Norm({ 123.0f, 114.0f, 89.0f }),</div><div class="line"><a name="l00453"></a><span class="lineno"> 453</span>  19.0f * CalcInvL2Norm({ 19.0f, 71.0f, 254.0f }),</div><div class="line"><a name="l00454"></a><span class="lineno"> 454</span>  172.0f * CalcInvL2Norm({ 172.0f, 122.0f, 12.0f }),</div><div class="line"><a name="l00455"></a><span class="lineno"> 455</span>  74.0f * CalcInvL2Norm({ 74.0f, 246.0f, 209.0f }),</div><div class="line"><a name="l00456"></a><span class="lineno"> 456</span>  250.0f * CalcInvL2Norm({ 250.0f, 166.0f, 200.0f }),</div><div class="line"><a name="l00457"></a><span class="lineno"> 457</span>  6.0f * CalcInvL2Norm({ 6.0f, 82.0f, 1.0f }),</div><div class="line"><a name="l00458"></a><span class="lineno"> 458</span>  195.0f * CalcInvL2Norm({ 195.0f, 28.0f, 64.0f }),</div><div class="line"><a name="l00459"></a><span class="lineno"> 459</span>  80.0f * CalcInvL2Norm({ 80.0f, 37.0f, 54.0f }),</div><div class="line"><a name="l00460"></a><span class="lineno"> 460</span> </div><div class="line"><a name="l00461"></a><span class="lineno"> 461</span>  <span class="comment">// Batch 0, Channel 1, Height (4) x Width (3)</span></div><div class="line"><a name="l00462"></a><span class="lineno"> 462</span>  113.0f * CalcInvL2Norm({ 235.0f, 113.0f, 56.0f }),</div><div class="line"><a name="l00463"></a><span class="lineno"> 463</span>  95.0f * CalcInvL2Norm({ 46.0f, 95.0f, 170.0f }),</div><div class="line"><a name="l00464"></a><span class="lineno"> 464</span>  202.0f * CalcInvL2Norm({ 178.0f, 202.0F, 162.0f }),</div><div class="line"><a name="l00465"></a><span class="lineno"> 465</span>  77.0f * CalcInvL2Norm({ 100.0f, 77.0f, 194.0f }),</div><div class="line"><a name="l00466"></a><span class="lineno"> 466</span>  114.0f * CalcInvL2Norm({ 123.0f, 114.0f, 89.0f }),</div><div class="line"><a name="l00467"></a><span class="lineno"> 467</span>  71.0f * CalcInvL2Norm({ 19.0f, 71.0f, 254.0f }),</div><div class="line"><a name="l00468"></a><span class="lineno"> 468</span>  122.0f * CalcInvL2Norm({ 172.0f, 122.0f, 12.0f }),</div><div class="line"><a name="l00469"></a><span class="lineno"> 469</span>  246.0f * CalcInvL2Norm({ 74.0f, 246.0f, 209.0f }),</div><div class="line"><a name="l00470"></a><span class="lineno"> 470</span>  166.0f * CalcInvL2Norm({ 250.0f, 166.0f, 200.0f }),</div><div class="line"><a name="l00471"></a><span class="lineno"> 471</span>  82.0f * CalcInvL2Norm({ 6.0f, 82.0f, 1.0f }),</div><div class="line"><a name="l00472"></a><span class="lineno"> 472</span>  28.0f * CalcInvL2Norm({ 195.0f, 28.0f, 64.0f }),</div><div class="line"><a name="l00473"></a><span class="lineno"> 473</span>  37.0f * CalcInvL2Norm({ 80.0f, 37.0f, 54.0f }),</div><div class="line"><a name="l00474"></a><span class="lineno"> 474</span> </div><div class="line"><a name="l00475"></a><span class="lineno"> 475</span>  <span class="comment">// Batch 0, Channel 2, Height (4) x Width (3)</span></div><div class="line"><a name="l00476"></a><span class="lineno"> 476</span>  56.0f * CalcInvL2Norm({ 235.0f, 113.0f, 56.0f }),</div><div class="line"><a name="l00477"></a><span class="lineno"> 477</span>  170.0f * CalcInvL2Norm({ 46.0f, 95.0f, 170.0f }),</div><div class="line"><a name="l00478"></a><span class="lineno"> 478</span>  162.0f * CalcInvL2Norm({ 178.0f, 202.0F, 162.0f }),</div><div class="line"><a name="l00479"></a><span class="lineno"> 479</span>  194.0f * CalcInvL2Norm({ 100.0f, 77.0f, 194.0f }),</div><div class="line"><a name="l00480"></a><span class="lineno"> 480</span>  89.0f * CalcInvL2Norm({ 123.0f, 114.0f, 89.0f }),</div><div class="line"><a name="l00481"></a><span class="lineno"> 481</span>  254.0f * CalcInvL2Norm({ 19.0f, 71.0f, 254.0f }),</div><div class="line"><a name="l00482"></a><span class="lineno"> 482</span>  12.0f * CalcInvL2Norm({ 172.0f, 122.0f, 12.0f }),</div><div class="line"><a name="l00483"></a><span class="lineno"> 483</span>  209.0f * CalcInvL2Norm({ 74.0f, 246.0f, 209.0f }),</div><div class="line"><a name="l00484"></a><span class="lineno"> 484</span>  200.0f * CalcInvL2Norm({ 250.0f, 166.0f, 200.0f }),</div><div class="line"><a name="l00485"></a><span class="lineno"> 485</span>  1.0f * CalcInvL2Norm({ 6.0f, 82.0f, 1.0f }),</div><div class="line"><a name="l00486"></a><span class="lineno"> 486</span>  64.0f * CalcInvL2Norm({ 195.0f, 28.0f, 64.0f }),</div><div class="line"><a name="l00487"></a><span class="lineno"> 487</span>  54.0f * CalcInvL2Norm({ 80.0f, 37.0f, 54.0f }),</div><div class="line"><a name="l00488"></a><span class="lineno"> 488</span> </div><div class="line"><a name="l00489"></a><span class="lineno"> 489</span>  <span class="comment">// Batch 1, Channel 0, Height (4) x Width (3)</span></div><div class="line"><a name="l00490"></a><span class="lineno"> 490</span>  67.0f * CalcInvL2Norm({ 67.0f, 239.0f, 97.0f }),</div><div class="line"><a name="l00491"></a><span class="lineno"> 491</span>  90.0f * CalcInvL2Norm({ 90.0f, 104.0f, 145.0f }),</div><div class="line"><a name="l00492"></a><span class="lineno"> 492</span>  49.0f * CalcInvL2Norm({ 49.0f, 199.0f, 215.0f }),</div><div class="line"><a name="l00493"></a><span class="lineno"> 493</span>  7.0f * CalcInvL2Norm({ 7.0f, 17.0f, 115.0f }),</div><div class="line"><a name="l00494"></a><span class="lineno"> 494</span>  163.0f * CalcInvL2Norm({ 163.0f, 124.0f, 116.0f }),</div><div class="line"><a name="l00495"></a><span class="lineno"> 495</span>  18.0f * CalcInvL2Norm({ 18.0f, 153.0f, 238.0f }),</div><div class="line"><a name="l00496"></a><span class="lineno"> 496</span>  25.0f * CalcInvL2Norm({ 25.0f, 222.0f, 226.0f }),</div><div class="line"><a name="l00497"></a><span class="lineno"> 497</span>  117.0f * CalcInvL2Norm({ 117.0f, 217.0f, 16.0f }),</div><div class="line"><a name="l00498"></a><span class="lineno"> 498</span>  103.0f * CalcInvL2Norm({ 103.0f, 75.0f, 132.0f }),</div><div class="line"><a name="l00499"></a><span class="lineno"> 499</span>  247.0f * CalcInvL2Norm({ 247.0f, 32.0f, 92.0f }),</div><div class="line"><a name="l00500"></a><span class="lineno"> 500</span>  59.0f * CalcInvL2Norm({ 59.0f, 126.0f, 125.0f }),</div><div class="line"><a name="l00501"></a><span class="lineno"> 501</span>  189.0f * CalcInvL2Norm({ 189.0f, 21.0f, 88.0f }),</div><div class="line"><a name="l00502"></a><span class="lineno"> 502</span> </div><div class="line"><a name="l00503"></a><span class="lineno"> 503</span>  <span class="comment">// Batch 1, Channel 1, Height (4) x Width (3)</span></div><div class="line"><a name="l00504"></a><span class="lineno"> 504</span>  239.0f * CalcInvL2Norm({ 67.0f, 239.0f, 97.0f }),</div><div class="line"><a name="l00505"></a><span class="lineno"> 505</span>  104.0f * CalcInvL2Norm({ 90.0f, 104.0f, 145.0f }),</div><div class="line"><a name="l00506"></a><span class="lineno"> 506</span>  199.0f * CalcInvL2Norm({ 49.0f, 199.0f, 215.0f }),</div><div class="line"><a name="l00507"></a><span class="lineno"> 507</span>  17.0f * CalcInvL2Norm({ 7.0f, 17.0f, 115.0f }),</div><div class="line"><a name="l00508"></a><span class="lineno"> 508</span>  124.0f * CalcInvL2Norm({ 163.0f, 124.0f, 116.0f }),</div><div class="line"><a name="l00509"></a><span class="lineno"> 509</span>  153.0f * CalcInvL2Norm({ 18.0f, 153.0f, 238.0f }),</div><div class="line"><a name="l00510"></a><span class="lineno"> 510</span>  222.0f * CalcInvL2Norm({ 25.0f, 222.0f, 226.0f }),</div><div class="line"><a name="l00511"></a><span class="lineno"> 511</span>  217.0f * CalcInvL2Norm({ 117.0f, 217.0f, 16.0f }),</div><div class="line"><a name="l00512"></a><span class="lineno"> 512</span>  75.0f * CalcInvL2Norm({ 103.0f, 75.0f, 132.0f }),</div><div class="line"><a name="l00513"></a><span class="lineno"> 513</span>  32.0f * CalcInvL2Norm({ 247.0f, 32.0f, 92.0f }),</div><div class="line"><a name="l00514"></a><span class="lineno"> 514</span>  126.0f * CalcInvL2Norm({ 59.0f, 126.0f, 125.0f }),</div><div class="line"><a name="l00515"></a><span class="lineno"> 515</span>  21.0f * CalcInvL2Norm({ 189.0f, 21.0f, 88.0f }),</div><div class="line"><a name="l00516"></a><span class="lineno"> 516</span> </div><div class="line"><a name="l00517"></a><span class="lineno"> 517</span>  <span class="comment">// Batch 1, Channel 2, Height (4) x Width (3)</span></div><div class="line"><a name="l00518"></a><span class="lineno"> 518</span>  97.0f * CalcInvL2Norm({ 67.0f, 239.0f, 97.0f }),</div><div class="line"><a name="l00519"></a><span class="lineno"> 519</span>  145.0f * CalcInvL2Norm({ 90.0f, 104.0f, 145.0f }),</div><div class="line"><a name="l00520"></a><span class="lineno"> 520</span>  215.0f * CalcInvL2Norm({ 49.0f, 199.0f, 215.0f }),</div><div class="line"><a name="l00521"></a><span class="lineno"> 521</span>  115.0f * CalcInvL2Norm({ 7.0f, 17.0f, 115.0f }),</div><div class="line"><a name="l00522"></a><span class="lineno"> 522</span>  116.0f * CalcInvL2Norm({ 163.0f, 124.0f, 116.0f }),</div><div class="line"><a name="l00523"></a><span class="lineno"> 523</span>  238.0f * CalcInvL2Norm({ 18.0f, 153.0f, 238.0f }),</div><div class="line"><a name="l00524"></a><span class="lineno"> 524</span>  226.0f * CalcInvL2Norm({ 25.0f, 222.0f, 226.0f }),</div><div class="line"><a name="l00525"></a><span class="lineno"> 525</span>  16.0f * CalcInvL2Norm({ 117.0f, 217.0f, 16.0f }),</div><div class="line"><a name="l00526"></a><span class="lineno"> 526</span>  132.0f * CalcInvL2Norm({ 103.0f, 75.0f, 132.0f }),</div><div class="line"><a name="l00527"></a><span class="lineno"> 527</span>  92.0f * CalcInvL2Norm({ 247.0f, 32.0f, 92.0f }),</div><div class="line"><a name="l00528"></a><span class="lineno"> 528</span>  125.0f * CalcInvL2Norm({ 59.0f, 126.0f, 125.0f }),</div><div class="line"><a name="l00529"></a><span class="lineno"> 529</span>  88.0f * CalcInvL2Norm({ 189.0f, 21.0f, 88.0f })</div><div class="line"><a name="l00530"></a><span class="lineno"> 530</span>  };</div><div class="line"><a name="l00531"></a><span class="lineno"> 531</span> </div><div class="line"><a name="l00532"></a><span class="lineno"> 532</span>  <span class="keywordflow">return</span> L2NormalizationTestImpl<ArmnnType>(</div><div class="line"><a name="l00533"></a><span class="lineno"> 533</span>  workloadFactory,</div><div class="line"><a name="l00534"></a><span class="lineno"> 534</span>  memoryManager,</div><div class="line"><a name="l00535"></a><span class="lineno"> 535</span>  inputOutputShape,</div><div class="line"><a name="l00536"></a><span class="lineno"> 536</span>  scale,</div><div class="line"><a name="l00537"></a><span class="lineno"> 537</span>  offset,</div><div class="line"><a name="l00538"></a><span class="lineno"> 538</span>  inputValues,</div><div class="line"><a name="l00539"></a><span class="lineno"> 539</span>  outScale,</div><div class="line"><a name="l00540"></a><span class="lineno"> 540</span>  outOffset,</div><div class="line"><a name="l00541"></a><span class="lineno"> 541</span>  expectedOutputValues,</div><div class="line"><a name="l00542"></a><span class="lineno"> 542</span>  layout);</div><div class="line"><a name="l00543"></a><span class="lineno"> 543</span> }</div><div class="line"><a name="l00544"></a><span class="lineno"> 544</span> </div><div class="line"><a name="l00545"></a><span class="lineno"> 545</span> } <span class="comment">// anonymous namespace</span></div><div class="line"><a name="l00546"></a><span class="lineno"> 546</span> </div><div class="line"><a name="l00547"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#a13c8cd6115422815348d57aef2ca032d"> 547</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<float, 4></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#a13c8cd6115422815348d57aef2ca032d">L2NormalizationDefaultEpsilonTest</a>(</div><div class="line"><a name="l00548"></a><span class="lineno"> 548</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00549"></a><span class="lineno"> 549</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00550"></a><span class="lineno"> 550</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00551"></a><span class="lineno"> 551</span> {</div><div class="line"><a name="l00552"></a><span class="lineno"> 552</span>  <span class="comment">// Dummy descriptor to get the default value of epsilon.</span></div><div class="line"><a name="l00553"></a><span class="lineno"> 553</span>  <a class="code" href="structarmnn_1_1_l2_normalization_descriptor.xhtml">armnn::L2NormalizationDescriptor</a> descriptor;</div><div class="line"><a name="l00554"></a><span class="lineno"> 554</span> </div><div class="line"><a name="l00555"></a><span class="lineno"> 555</span>  <span class="keywordflow">return</span> L2NormalizationEpsilonTestCommon<armnn::DataType::Float32>(</div><div class="line"><a name="l00556"></a><span class="lineno"> 556</span>  workloadFactory,</div><div class="line"><a name="l00557"></a><span class="lineno"> 557</span>  memoryManager,</div><div class="line"><a name="l00558"></a><span class="lineno"> 558</span>  0.f,</div><div class="line"><a name="l00559"></a><span class="lineno"> 559</span>  0,</div><div class="line"><a name="l00560"></a><span class="lineno"> 560</span>  0.f,</div><div class="line"><a name="l00561"></a><span class="lineno"> 561</span>  0,</div><div class="line"><a name="l00562"></a><span class="lineno"> 562</span>  layout,</div><div class="line"><a name="l00563"></a><span class="lineno"> 563</span>  descriptor.<a class="code" href="structarmnn_1_1_l2_normalization_descriptor.xhtml#a11c821c7524251004a72ed13c510853c">m_Eps</a>);</div><div class="line"><a name="l00564"></a><span class="lineno"> 564</span> }</div><div class="line"><a name="l00565"></a><span class="lineno"> 565</span> </div><div class="line"><a name="l00566"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#ae6ec1c0ad5b1b94d03c160c8122587cc"> 566</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<float, 4></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#ae6ec1c0ad5b1b94d03c160c8122587cc">L2NormalizationNonDefaultEpsilonTest</a>(</div><div class="line"><a name="l00567"></a><span class="lineno"> 567</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00568"></a><span class="lineno"> 568</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00569"></a><span class="lineno"> 569</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00570"></a><span class="lineno"> 570</span> {</div><div class="line"><a name="l00571"></a><span class="lineno"> 571</span>  <span class="keywordflow">return</span> L2NormalizationEpsilonTestCommon<armnn::DataType::Float32>(</div><div class="line"><a name="l00572"></a><span class="lineno"> 572</span>  workloadFactory,</div><div class="line"><a name="l00573"></a><span class="lineno"> 573</span>  memoryManager,</div><div class="line"><a name="l00574"></a><span class="lineno"> 574</span>  0.f,</div><div class="line"><a name="l00575"></a><span class="lineno"> 575</span>  0,</div><div class="line"><a name="l00576"></a><span class="lineno"> 576</span>  0.f,</div><div class="line"><a name="l00577"></a><span class="lineno"> 577</span>  0,</div><div class="line"><a name="l00578"></a><span class="lineno"> 578</span>  layout,</div><div class="line"><a name="l00579"></a><span class="lineno"> 579</span>  1e-9f);</div><div class="line"><a name="l00580"></a><span class="lineno"> 580</span> }</div><div class="line"><a name="l00581"></a><span class="lineno"> 581</span> </div><div class="line"><a name="l00582"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#abc9aa62ee9cdec8c43b5a43d931c632c"> 582</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<float, 4></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#abc9aa62ee9cdec8c43b5a43d931c632c">L2Normalization1dTest</a>(</div><div class="line"><a name="l00583"></a><span class="lineno"> 583</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00584"></a><span class="lineno"> 584</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00585"></a><span class="lineno"> 585</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00586"></a><span class="lineno"> 586</span> {</div><div class="line"><a name="l00587"></a><span class="lineno"> 587</span>  <span class="keywordflow">return</span> L2Normalization1dTestCommon<armnn::DataType::Float32>(</div><div class="line"><a name="l00588"></a><span class="lineno"> 588</span>  workloadFactory,</div><div class="line"><a name="l00589"></a><span class="lineno"> 589</span>  memoryManager,</div><div class="line"><a name="l00590"></a><span class="lineno"> 590</span>  0.f,</div><div class="line"><a name="l00591"></a><span class="lineno"> 591</span>  0,</div><div class="line"><a name="l00592"></a><span class="lineno"> 592</span>  0.f,</div><div class="line"><a name="l00593"></a><span class="lineno"> 593</span>  0,</div><div class="line"><a name="l00594"></a><span class="lineno"> 594</span>  layout);</div><div class="line"><a name="l00595"></a><span class="lineno"> 595</span> }</div><div class="line"><a name="l00596"></a><span class="lineno"> 596</span> </div><div class="line"><a name="l00597"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#a51324dd32b0b605e9f27d2b91312dc80"> 597</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<int16_t, 4></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#a51324dd32b0b605e9f27d2b91312dc80">L2Normalization1dInt16Test</a>(</div><div class="line"><a name="l00598"></a><span class="lineno"> 598</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00599"></a><span class="lineno"> 599</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00600"></a><span class="lineno"> 600</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00601"></a><span class="lineno"> 601</span> {</div><div class="line"><a name="l00602"></a><span class="lineno"> 602</span>  <span class="keywordflow">return</span> L2Normalization1dTestCommon<armnn::DataType::QSymmS16>(</div><div class="line"><a name="l00603"></a><span class="lineno"> 603</span>  workloadFactory,</div><div class="line"><a name="l00604"></a><span class="lineno"> 604</span>  memoryManager,</div><div class="line"><a name="l00605"></a><span class="lineno"> 605</span>  1.f,</div><div class="line"><a name="l00606"></a><span class="lineno"> 606</span>  0,</div><div class="line"><a name="l00607"></a><span class="lineno"> 607</span>  1.f,</div><div class="line"><a name="l00608"></a><span class="lineno"> 608</span>  0,</div><div class="line"><a name="l00609"></a><span class="lineno"> 609</span>  layout);</div><div class="line"><a name="l00610"></a><span class="lineno"> 610</span> }</div><div class="line"><a name="l00611"></a><span class="lineno"> 611</span> </div><div class="line"><a name="l00612"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#aea6a1743ba0fdb000d73856302ab6c23"> 612</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<uint8_t, 4></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#aea6a1743ba0fdb000d73856302ab6c23">L2Normalization1dUint8Test</a>(</div><div class="line"><a name="l00613"></a><span class="lineno"> 613</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00614"></a><span class="lineno"> 614</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00615"></a><span class="lineno"> 615</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00616"></a><span class="lineno"> 616</span> {</div><div class="line"><a name="l00617"></a><span class="lineno"> 617</span>  <span class="keywordflow">return</span> L2Normalization1dTestCommon<armnn::DataType::QAsymmU8>(</div><div class="line"><a name="l00618"></a><span class="lineno"> 618</span>  workloadFactory,</div><div class="line"><a name="l00619"></a><span class="lineno"> 619</span>  memoryManager,</div><div class="line"><a name="l00620"></a><span class="lineno"> 620</span>  1.f,</div><div class="line"><a name="l00621"></a><span class="lineno"> 621</span>  0,</div><div class="line"><a name="l00622"></a><span class="lineno"> 622</span>  1.f / 128,</div><div class="line"><a name="l00623"></a><span class="lineno"> 623</span>  128,</div><div class="line"><a name="l00624"></a><span class="lineno"> 624</span>  layout);</div><div class="line"><a name="l00625"></a><span class="lineno"> 625</span> }</div><div class="line"><a name="l00626"></a><span class="lineno"> 626</span> </div><div class="line"><a name="l00627"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#a373fc44a34b2bba8739ad4c6e864b234"> 627</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<float, 4></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#a373fc44a34b2bba8739ad4c6e864b234">L2Normalization2dTest</a>(</div><div class="line"><a name="l00628"></a><span class="lineno"> 628</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00629"></a><span class="lineno"> 629</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00630"></a><span class="lineno"> 630</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00631"></a><span class="lineno"> 631</span> {</div><div class="line"><a name="l00632"></a><span class="lineno"> 632</span>  <span class="keywordflow">return</span> L2Normalization2dTestCommon<armnn::DataType::Float32>(</div><div class="line"><a name="l00633"></a><span class="lineno"> 633</span>  workloadFactory,</div><div class="line"><a name="l00634"></a><span class="lineno"> 634</span>  memoryManager,</div><div class="line"><a name="l00635"></a><span class="lineno"> 635</span>  0.f,</div><div class="line"><a name="l00636"></a><span class="lineno"> 636</span>  0,</div><div class="line"><a name="l00637"></a><span class="lineno"> 637</span>  0.f,</div><div class="line"><a name="l00638"></a><span class="lineno"> 638</span>  0,</div><div class="line"><a name="l00639"></a><span class="lineno"> 639</span>  layout);</div><div class="line"><a name="l00640"></a><span class="lineno"> 640</span> }</div><div class="line"><a name="l00641"></a><span class="lineno"> 641</span> </div><div class="line"><a name="l00642"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#a6e2879224854a663f502d3092a68d2c7"> 642</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<int16_t, 4></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#a6e2879224854a663f502d3092a68d2c7">L2Normalization2dInt16Test</a>(</div><div class="line"><a name="l00643"></a><span class="lineno"> 643</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00644"></a><span class="lineno"> 644</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00645"></a><span class="lineno"> 645</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00646"></a><span class="lineno"> 646</span> {</div><div class="line"><a name="l00647"></a><span class="lineno"> 647</span>  <span class="keywordflow">return</span> L2Normalization1dTestCommon<armnn::DataType::QSymmS16>(</div><div class="line"><a name="l00648"></a><span class="lineno"> 648</span>  workloadFactory,</div><div class="line"><a name="l00649"></a><span class="lineno"> 649</span>  memoryManager,</div><div class="line"><a name="l00650"></a><span class="lineno"> 650</span>  1.f,</div><div class="line"><a name="l00651"></a><span class="lineno"> 651</span>  0,</div><div class="line"><a name="l00652"></a><span class="lineno"> 652</span>  1.f,</div><div class="line"><a name="l00653"></a><span class="lineno"> 653</span>  0,</div><div class="line"><a name="l00654"></a><span class="lineno"> 654</span>  layout);</div><div class="line"><a name="l00655"></a><span class="lineno"> 655</span> }</div><div class="line"><a name="l00656"></a><span class="lineno"> 656</span> </div><div class="line"><a name="l00657"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#a96cf65cb33a0e9319ddd0d00d56b5056"> 657</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<uint8_t, 4></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#a96cf65cb33a0e9319ddd0d00d56b5056">L2Normalization2dUint8Test</a>(</div><div class="line"><a name="l00658"></a><span class="lineno"> 658</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00659"></a><span class="lineno"> 659</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00660"></a><span class="lineno"> 660</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00661"></a><span class="lineno"> 661</span> {</div><div class="line"><a name="l00662"></a><span class="lineno"> 662</span>  <span class="keywordflow">return</span> L2Normalization1dTestCommon<armnn::DataType::QAsymmU8>(</div><div class="line"><a name="l00663"></a><span class="lineno"> 663</span>  workloadFactory,</div><div class="line"><a name="l00664"></a><span class="lineno"> 664</span>  memoryManager,</div><div class="line"><a name="l00665"></a><span class="lineno"> 665</span>  1.f,</div><div class="line"><a name="l00666"></a><span class="lineno"> 666</span>  0,</div><div class="line"><a name="l00667"></a><span class="lineno"> 667</span>  1.f / 128,</div><div class="line"><a name="l00668"></a><span class="lineno"> 668</span>  128,</div><div class="line"><a name="l00669"></a><span class="lineno"> 669</span>  layout);</div><div class="line"><a name="l00670"></a><span class="lineno"> 670</span> }</div><div class="line"><a name="l00671"></a><span class="lineno"> 671</span> </div><div class="line"><a name="l00672"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#a1a90f524b460439bb8e386ea672acd6c"> 672</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<float, 2></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#a1a90f524b460439bb8e386ea672acd6c">L2Normalization2dShapeTest</a>(</div><div class="line"><a name="l00673"></a><span class="lineno"> 673</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00674"></a><span class="lineno"> 674</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager)</div><div class="line"><a name="l00675"></a><span class="lineno"> 675</span> {</div><div class="line"><a name="l00676"></a><span class="lineno"> 676</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout = <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0ad066db54b89b0912e7e7c6da51e2da51">armnn::DataLayout::NHWC</a>;</div><div class="line"><a name="l00677"></a><span class="lineno"> 677</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_shape.xhtml">armnn::TensorShape</a> inputOutputTensorShape = <a class="code" href="classarmnn_1_1_tensor_shape.xhtml">armnn::TensorShape</a>({ 5, 2 });</div><div class="line"><a name="l00678"></a><span class="lineno"> 678</span> </div><div class="line"><a name="l00679"></a><span class="lineno"> 679</span>  std::vector<float> inputData</div><div class="line"><a name="l00680"></a><span class="lineno"> 680</span>  {</div><div class="line"><a name="l00681"></a><span class="lineno"> 681</span>  1.f, 2.f, 3.f, 4.f, 5.f, 6.f, 7.f, 8.f, 9.f, 10.f</div><div class="line"><a name="l00682"></a><span class="lineno"> 682</span>  };</div><div class="line"><a name="l00683"></a><span class="lineno"> 683</span>  std::vector<float> expectedOutputData</div><div class="line"><a name="l00684"></a><span class="lineno"> 684</span>  {</div><div class="line"><a name="l00685"></a><span class="lineno"> 685</span>  1.0f * CalcInvL2Norm({ 1.0f, 2.0f }),</div><div class="line"><a name="l00686"></a><span class="lineno"> 686</span>  2.0f * CalcInvL2Norm({ 1.0f, 2.0f }),</div><div class="line"><a name="l00687"></a><span class="lineno"> 687</span>  3.0f * CalcInvL2Norm({ 3.0f, 4.0f }),</div><div class="line"><a name="l00688"></a><span class="lineno"> 688</span>  4.0f * CalcInvL2Norm({ 3.0f, 4.0f }),</div><div class="line"><a name="l00689"></a><span class="lineno"> 689</span>  5.0f * CalcInvL2Norm({ 5.0f, 6.0f }),</div><div class="line"><a name="l00690"></a><span class="lineno"> 690</span>  6.0f * CalcInvL2Norm({ 5.0f, 6.0f }),</div><div class="line"><a name="l00691"></a><span class="lineno"> 691</span>  7.0f * CalcInvL2Norm({ 7.0f, 8.0f }),</div><div class="line"><a name="l00692"></a><span class="lineno"> 692</span>  8.0f * CalcInvL2Norm({ 7.0f, 8.0f }),</div><div class="line"><a name="l00693"></a><span class="lineno"> 693</span>  9.0f * CalcInvL2Norm({ 9.0f, 10.0f }),</div><div class="line"><a name="l00694"></a><span class="lineno"> 694</span>  10.0f * CalcInvL2Norm({ 9.0f, 10.0f })</div><div class="line"><a name="l00695"></a><span class="lineno"> 695</span>  };</div><div class="line"><a name="l00696"></a><span class="lineno"> 696</span> </div><div class="line"><a name="l00697"></a><span class="lineno"> 697</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.xhtml">armnn::TensorInfo</a> inputTensorInfo(inputOutputTensorShape, <a class="code" href="namespacearmnn.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">armnn::DataType::Float32</a>, 0.f, 0);</div><div class="line"><a name="l00698"></a><span class="lineno"> 698</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_tensor_info.xhtml">armnn::TensorInfo</a> outputTensorInfo(inputOutputTensorShape, <a class="code" href="namespacearmnn.xhtml#ad8ed01ff3ff33333d8e19db4d2818bb6a166495adc0d0f53bee6baecc577f5204">armnn::DataType::Float32</a>, 0.f, 0);</div><div class="line"><a name="l00699"></a><span class="lineno"> 699</span> </div><div class="line"><a name="l00700"></a><span class="lineno"> 700</span>  <span class="keyword">auto</span> inputTensor = MakeTensor<float, 2>(inputTensorInfo, inputData);</div><div class="line"><a name="l00701"></a><span class="lineno"> 701</span> </div><div class="line"><a name="l00702"></a><span class="lineno"> 702</span>  <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<float, 2></a> result(outputTensorInfo);</div><div class="line"><a name="l00703"></a><span class="lineno"> 703</span>  result.outputExpected = MakeTensor<float, 2>(outputTensorInfo, expectedOutputData);</div><div class="line"><a name="l00704"></a><span class="lineno"> 704</span> </div><div class="line"><a name="l00705"></a><span class="lineno"> 705</span>  std::unique_ptr<armnn::ITensorHandle> inputHandle = workloadFactory.<a class="code" href="classarmnn_1_1_i_workload_factory.xhtml#a15c140be4ddceffee16436f009d3ed94">CreateTensorHandle</a>(inputTensorInfo);</div><div class="line"><a name="l00706"></a><span class="lineno"> 706</span>  std::unique_ptr<armnn::ITensorHandle> outputHandle = workloadFactory.<a class="code" href="classarmnn_1_1_i_workload_factory.xhtml#a15c140be4ddceffee16436f009d3ed94">CreateTensorHandle</a>(outputTensorInfo);</div><div class="line"><a name="l00707"></a><span class="lineno"> 707</span> </div><div class="line"><a name="l00708"></a><span class="lineno"> 708</span>  <a class="code" href="structarmnn_1_1_l2_normalization_queue_descriptor.xhtml">armnn::L2NormalizationQueueDescriptor</a> descriptor;</div><div class="line"><a name="l00709"></a><span class="lineno"> 709</span>  descriptor.<a class="code" href="structarmnn_1_1_queue_descriptor_with_parameters.xhtml#aad91b9bbf7aa365d304febe79a3d1333">m_Parameters</a>.<a class="code" href="structarmnn_1_1_l2_normalization_descriptor.xhtml#a11c821c7524251004a72ed13c510853c">m_Eps</a> = 1e-12f;</div><div class="line"><a name="l00710"></a><span class="lineno"> 710</span>  descriptor.<a class="code" href="structarmnn_1_1_queue_descriptor_with_parameters.xhtml#aad91b9bbf7aa365d304febe79a3d1333">m_Parameters</a>.<a class="code" href="structarmnn_1_1_l2_normalization_descriptor.xhtml#a6089e1ca91914015777ea780a513131a">m_DataLayout</a> = layout;</div><div class="line"><a name="l00711"></a><span class="lineno"> 711</span>  <a class="code" href="structarmnn_1_1_workload_info.xhtml">armnn::WorkloadInfo</a> info;</div><div class="line"><a name="l00712"></a><span class="lineno"> 712</span> </div><div class="line"><a name="l00713"></a><span class="lineno"> 713</span>  AddInputToWorkload(descriptor, info, inputTensorInfo, inputHandle.get());</div><div class="line"><a name="l00714"></a><span class="lineno"> 714</span>  AddOutputToWorkload(descriptor, info, outputTensorInfo, outputHandle.get());</div><div class="line"><a name="l00715"></a><span class="lineno"> 715</span> </div><div class="line"><a name="l00716"></a><span class="lineno"> 716</span>  std::unique_ptr<armnn::IWorkload> workload = workloadFactory.<a class="code" href="classarmnn_1_1_i_workload_factory.xhtml#a3c86f886e36ce943f1ebc241a37f0413">CreateL2Normalization</a>(descriptor, info);</div><div class="line"><a name="l00717"></a><span class="lineno"> 717</span> </div><div class="line"><a name="l00718"></a><span class="lineno"> 718</span>  inputHandle->Allocate();</div><div class="line"><a name="l00719"></a><span class="lineno"> 719</span>  outputHandle->Allocate();</div><div class="line"><a name="l00720"></a><span class="lineno"> 720</span> </div><div class="line"><a name="l00721"></a><span class="lineno"> 721</span>  <a class="code" href="_tensor_copy_utils_8cpp.xhtml#ae15f1a3c55d2db87683577de9fa4437c">CopyDataToITensorHandle</a>(inputHandle.get(), &inputTensor[0][0]);</div><div class="line"><a name="l00722"></a><span class="lineno"> 722</span> </div><div class="line"><a name="l00723"></a><span class="lineno"> 723</span>  workload->PostAllocationConfigure();</div><div class="line"><a name="l00724"></a><span class="lineno"> 724</span>  ExecuteWorkload(*workload, memoryManager);</div><div class="line"><a name="l00725"></a><span class="lineno"> 725</span> </div><div class="line"><a name="l00726"></a><span class="lineno"> 726</span>  <a class="code" href="_tensor_copy_utils_8cpp.xhtml#a99b626c58a926dc7d6df78d22ec186c8">CopyDataFromITensorHandle</a>(&result.output[0][0], outputHandle.get());</div><div class="line"><a name="l00727"></a><span class="lineno"> 727</span> </div><div class="line"><a name="l00728"></a><span class="lineno"> 728</span>  <span class="keywordflow">return</span> result;</div><div class="line"><a name="l00729"></a><span class="lineno"> 729</span> }</div><div class="line"><a name="l00730"></a><span class="lineno"> 730</span> </div><div class="line"><a name="l00731"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#a336e63cb246a1d6f8b5a02367932471a"> 731</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<float, 4></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#a336e63cb246a1d6f8b5a02367932471a">L2Normalization3dTest</a>(</div><div class="line"><a name="l00732"></a><span class="lineno"> 732</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00733"></a><span class="lineno"> 733</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00734"></a><span class="lineno"> 734</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00735"></a><span class="lineno"> 735</span> {</div><div class="line"><a name="l00736"></a><span class="lineno"> 736</span>  <span class="keywordflow">return</span> L2Normalization3dTestCommon<armnn::DataType::Float32>(</div><div class="line"><a name="l00737"></a><span class="lineno"> 737</span>  workloadFactory,</div><div class="line"><a name="l00738"></a><span class="lineno"> 738</span>  memoryManager,</div><div class="line"><a name="l00739"></a><span class="lineno"> 739</span>  0.f,</div><div class="line"><a name="l00740"></a><span class="lineno"> 740</span>  0,</div><div class="line"><a name="l00741"></a><span class="lineno"> 741</span>  0.f,</div><div class="line"><a name="l00742"></a><span class="lineno"> 742</span>  0,</div><div class="line"><a name="l00743"></a><span class="lineno"> 743</span>  layout);</div><div class="line"><a name="l00744"></a><span class="lineno"> 744</span> }</div><div class="line"><a name="l00745"></a><span class="lineno"> 745</span> </div><div class="line"><a name="l00746"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#a49295d2552ff6a80396649f5b6e3a9ce"> 746</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<int16_t, 4></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#a49295d2552ff6a80396649f5b6e3a9ce">L2Normalization3dInt16Test</a>(</div><div class="line"><a name="l00747"></a><span class="lineno"> 747</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00748"></a><span class="lineno"> 748</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00749"></a><span class="lineno"> 749</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00750"></a><span class="lineno"> 750</span> {</div><div class="line"><a name="l00751"></a><span class="lineno"> 751</span>  <span class="keywordflow">return</span> L2Normalization1dTestCommon<armnn::DataType::QSymmS16>(</div><div class="line"><a name="l00752"></a><span class="lineno"> 752</span>  workloadFactory,</div><div class="line"><a name="l00753"></a><span class="lineno"> 753</span>  memoryManager,</div><div class="line"><a name="l00754"></a><span class="lineno"> 754</span>  1.f,</div><div class="line"><a name="l00755"></a><span class="lineno"> 755</span>  0,</div><div class="line"><a name="l00756"></a><span class="lineno"> 756</span>  1.f,</div><div class="line"><a name="l00757"></a><span class="lineno"> 757</span>  0,</div><div class="line"><a name="l00758"></a><span class="lineno"> 758</span>  layout);</div><div class="line"><a name="l00759"></a><span class="lineno"> 759</span> }</div><div class="line"><a name="l00760"></a><span class="lineno"> 760</span> </div><div class="line"><a name="l00761"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#ad087db636160f71155a4ac31b37184aa"> 761</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<uint8_t, 4></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#ad087db636160f71155a4ac31b37184aa">L2Normalization3dUint8Test</a>(</div><div class="line"><a name="l00762"></a><span class="lineno"> 762</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00763"></a><span class="lineno"> 763</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00764"></a><span class="lineno"> 764</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00765"></a><span class="lineno"> 765</span> {</div><div class="line"><a name="l00766"></a><span class="lineno"> 766</span>  <span class="keywordflow">return</span> L2Normalization1dTestCommon<armnn::DataType::QAsymmU8>(</div><div class="line"><a name="l00767"></a><span class="lineno"> 767</span>  workloadFactory,</div><div class="line"><a name="l00768"></a><span class="lineno"> 768</span>  memoryManager,</div><div class="line"><a name="l00769"></a><span class="lineno"> 769</span>  1.f,</div><div class="line"><a name="l00770"></a><span class="lineno"> 770</span>  0,</div><div class="line"><a name="l00771"></a><span class="lineno"> 771</span>  1.f / 128,</div><div class="line"><a name="l00772"></a><span class="lineno"> 772</span>  128,</div><div class="line"><a name="l00773"></a><span class="lineno"> 773</span>  layout);</div><div class="line"><a name="l00774"></a><span class="lineno"> 774</span> }</div><div class="line"><a name="l00775"></a><span class="lineno"> 775</span> </div><div class="line"><a name="l00776"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#a49c8b464589cbe8f6b7e7a1bf7e6403d"> 776</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<float, 4></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#a49c8b464589cbe8f6b7e7a1bf7e6403d">L2Normalization4dTest</a>(</div><div class="line"><a name="l00777"></a><span class="lineno"> 777</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00778"></a><span class="lineno"> 778</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00779"></a><span class="lineno"> 779</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00780"></a><span class="lineno"> 780</span> {</div><div class="line"><a name="l00781"></a><span class="lineno"> 781</span>  <span class="keywordflow">return</span> L2Normalization4dTestCommon<armnn::DataType::Float32>(</div><div class="line"><a name="l00782"></a><span class="lineno"> 782</span>  workloadFactory,</div><div class="line"><a name="l00783"></a><span class="lineno"> 783</span>  memoryManager,</div><div class="line"><a name="l00784"></a><span class="lineno"> 784</span>  0.f,</div><div class="line"><a name="l00785"></a><span class="lineno"> 785</span>  0,</div><div class="line"><a name="l00786"></a><span class="lineno"> 786</span>  0.f,</div><div class="line"><a name="l00787"></a><span class="lineno"> 787</span>  0,</div><div class="line"><a name="l00788"></a><span class="lineno"> 788</span>  layout);</div><div class="line"><a name="l00789"></a><span class="lineno"> 789</span> }</div><div class="line"><a name="l00790"></a><span class="lineno"> 790</span> </div><div class="line"><a name="l00791"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#a1bbff607f046d49a92516969d8beff7a"> 791</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<int16_t, 4></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#a1bbff607f046d49a92516969d8beff7a">L2Normalization4dInt16Test</a>(</div><div class="line"><a name="l00792"></a><span class="lineno"> 792</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00793"></a><span class="lineno"> 793</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00794"></a><span class="lineno"> 794</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00795"></a><span class="lineno"> 795</span> {</div><div class="line"><a name="l00796"></a><span class="lineno"> 796</span>  <span class="keywordflow">return</span> L2Normalization1dTestCommon<armnn::DataType::QSymmS16>(</div><div class="line"><a name="l00797"></a><span class="lineno"> 797</span>  workloadFactory,</div><div class="line"><a name="l00798"></a><span class="lineno"> 798</span>  memoryManager,</div><div class="line"><a name="l00799"></a><span class="lineno"> 799</span>  1.f,</div><div class="line"><a name="l00800"></a><span class="lineno"> 800</span>  0,</div><div class="line"><a name="l00801"></a><span class="lineno"> 801</span>  1.f,</div><div class="line"><a name="l00802"></a><span class="lineno"> 802</span>  0,</div><div class="line"><a name="l00803"></a><span class="lineno"> 803</span>  layout);</div><div class="line"><a name="l00804"></a><span class="lineno"> 804</span> }</div><div class="line"><a name="l00805"></a><span class="lineno"> 805</span> </div><div class="line"><a name="l00806"></a><span class="lineno"><a class="line" href="_l2_normalization_test_impl_8hpp.xhtml#a9672054d1096864d4c034aa90008efff"> 806</a></span> <a class="code" href="struct_layer_test_result.xhtml">LayerTestResult<uint8_t, 4></a> <a class="code" href="_l2_normalization_test_impl_8cpp.xhtml#a9672054d1096864d4c034aa90008efff">L2Normalization4dUint8Test</a>(</div><div class="line"><a name="l00807"></a><span class="lineno"> 807</span>  <a class="code" href="classarmnn_1_1_i_workload_factory.xhtml">armnn::IWorkloadFactory</a>& workloadFactory,</div><div class="line"><a name="l00808"></a><span class="lineno"> 808</span>  <span class="keyword">const</span> <a class="code" href="classarmnn_1_1_i_backend_internal.xhtml#a693b40e6b94e958836aeb0410ca186bd">armnn::IBackendInternal::IMemoryManagerSharedPtr</a>& memoryManager,</div><div class="line"><a name="l00809"></a><span class="lineno"> 809</span>  <span class="keyword">const</span> <a class="code" href="namespacearmnn.xhtml#ad1d5cce2d9e9a5d61c243e5c989112e0">armnn::DataLayout</a> layout)</div><div class="line"><a name="l00810"></a><span class="lineno"> 810</span> {</div><div class="line"><a name="l00811"></a><span class="lineno"> 811</span>  <span class="keywordflow">return</span> L2Normalization1dTestCommon<armnn::DataType::QAsymmU8>(</div><div class="line"><a name="l00812"></a><span class="lineno"> 812</span>  workloadFactory,</div><div class="line"><a name="l00813"></a><span class="lineno"> 813</span>  memoryManager,</div><div class="line"><a name="l00814"></a><span class="lineno"> 814</span>  1.f,</div><div class="line"><a name="l00815"></a><span class="lineno"> 815</span>  0,</div><div class="line"><a name="l00816"></a><span class="lineno"> 816</span>  1.f / 128,</div><div class="line"><a name="l00817"></a><span class="lineno"> 817</span>  128,</div><div class="line"><a name="l00818"></a><span class="lineno"> 818</span>  layout);</div><div class="line"><a name="l00819"></a><span class="lineno"> 819</span> }</div><div class="ttc" id="structarmnn_1_1_l2_normalization_descriptor_xhtml_a11c821c7524251004a72ed13c510853c"><div class="ttname"><a href="structarmnn_1_1_l2_normalization_descriptor.xhtml#a11c821c7524251004a72ed13c510853c">armnn::L2NormalizationDescriptor::m_Eps</a></div><div class="ttdeci">float m_Eps</div><div class="ttdoc">Used to avoid dividing by zero. </div><div class="ttdef"><b>Definition:</b> <a href="_descriptors_8hpp_source.xhtml#l00604">Descriptors.hpp:604</a></div></div> |