| <!-- HTML header for doxygen 1.8.15--> |
| <!-- Remember to use version doxygen 1.8.15 +--> |
| <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> |
| <html xmlns="http://www.w3.org/1999/xhtml"> |
| <head> |
| <meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/> |
| <meta http-equiv="X-UA-Compatible" content="IE=9"/> |
| <meta name="generator" content="Doxygen 1.8.15"/> |
| <meta name="robots" content="NOINDEX, NOFOLLOW" /> <!-- Prevent indexing by search engines --> |
| <title>Compute Library: arm_compute/runtime/CL/functions/CLGEMMLowpOutputStage.h Source File</title> |
| <link href="tabs.css" rel="stylesheet" type="text/css"/> |
| <script type="text/javascript" src="jquery.js"></script> |
| <script type="text/javascript" src="dynsections.js"></script> |
| <link href="navtree.css" rel="stylesheet" type="text/css"/> |
| <script type="text/javascript" src="resize.js"></script> |
| <script type="text/javascript" src="navtreedata.js"></script> |
| <script type="text/javascript" src="navtree.js"></script> |
| <script type="text/javascript"> |
| /* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&dn=gpl-2.0.txt GPL-v2 */ |
| $(document).ready(initResizable); |
| /* @license-end */</script> |
| <link href="search/search.css" rel="stylesheet" type="text/css"/> |
| <script type="text/javascript" src="search/searchdata.js"></script> |
| <script type="text/javascript" src="search/search.js"></script> |
| <script type="text/x-mathjax-config"> |
| MathJax.Hub.Config({ |
| extensions: ["tex2jax.js"], |
| jax: ["input/TeX","output/HTML-CSS"], |
| }); |
| </script><script type="text/javascript" async="async" src="http://cdn.mathjax.org/mathjax/latest/MathJax.js"></script> |
| <link href="doxygen.css" rel="stylesheet" type="text/css" /> |
| <link href="stylesheet.css" rel="stylesheet" type="text/css"/> |
| </head> |
| <body> |
| <div id="top"><!-- do not remove this div, it is closed by doxygen! --> |
| <div id="titlearea"> |
| <table cellspacing="0" cellpadding="0"> |
| <tbody> |
| <tr style="height: 56px;"> |
| <img alt="Compute Library" src="https://raw.githubusercontent.com/ARM-software/ComputeLibrary/gh-pages/ACL_logo.png" style="max-width: 100%;margin-top: 15px;margin-left: 10px"/> |
| <td style="padding-left: 0.5em;"> |
| <div id="projectname"> |
|  <span id="projectnumber">20.02.1</span> |
| </div> |
| </td> |
| </tr> |
| </tbody> |
| </table> |
| </div> |
| <!-- end header part --> |
| <!-- Generated by Doxygen 1.8.15 --> |
| <script type="text/javascript"> |
| /* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&dn=gpl-2.0.txt GPL-v2 */ |
| var searchBox = new SearchBox("searchBox", "search",false,'Search'); |
| /* @license-end */ |
| </script> |
| <script type="text/javascript" src="menudata.js"></script> |
| <script type="text/javascript" src="menu.js"></script> |
| <script type="text/javascript"> |
| /* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&dn=gpl-2.0.txt GPL-v2 */ |
| $(function() { |
| initMenu('',true,false,'search.php','Search'); |
| $(document).ready(function() { init_search(); }); |
| }); |
| /* @license-end */</script> |
| <div id="main-nav"></div> |
| </div><!-- top --> |
| <div id="side-nav" class="ui-resizable side-nav-resizable"> |
| <div id="nav-tree"> |
| <div id="nav-tree-contents"> |
| <div id="nav-sync" class="sync"></div> |
| </div> |
| </div> |
| <div id="splitbar" style="-moz-user-select:none;" |
| class="ui-resizable-handle"> |
| </div> |
| </div> |
| <script type="text/javascript"> |
| /* @license magnet:?xt=urn:btih:cf05388f2679ee054f2beb29a391d25f4e673ac3&dn=gpl-2.0.txt GPL-v2 */ |
| $(document).ready(function(){initNavTree('_c_l_g_e_m_m_lowp_output_stage_8h_source.xhtml','');}); |
| /* @license-end */ |
| </script> |
| <div id="doc-content"> |
| <!-- window showing the filter options --> |
| <div id="MSearchSelectWindow" |
| onmouseover="return searchBox.OnSearchSelectShow()" |
| onmouseout="return searchBox.OnSearchSelectHide()" |
| onkeydown="return searchBox.OnSearchSelectKey(event)"> |
| </div> |
| |
| <!-- iframe showing the search results (closed by default) --> |
| <div id="MSearchResultsWindow"> |
| <iframe src="javascript:void(0)" frameborder="0" |
| name="MSearchResults" id="MSearchResults"> |
| </iframe> |
| </div> |
| |
| <div class="header"> |
| <div class="headertitle"> |
| <div class="title">CLGEMMLowpOutputStage.h</div> </div> |
| </div><!--header--> |
| <div class="contents"> |
| <a href="_c_l_g_e_m_m_lowp_output_stage_8h.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 (c) 2017-2019 ARM Limited.</span></div><div class="line"><a name="l00003"></a><span class="lineno"> 3</span> <span class="comment"> *</span></div><div class="line"><a name="l00004"></a><span class="lineno"> 4</span> <span class="comment"> * SPDX-License-Identifier: MIT</span></div><div class="line"><a name="l00005"></a><span class="lineno"> 5</span> <span class="comment"> *</span></div><div class="line"><a name="l00006"></a><span class="lineno"> 6</span> <span class="comment"> * Permission is hereby granted, free of charge, to any person obtaining a copy</span></div><div class="line"><a name="l00007"></a><span class="lineno"> 7</span> <span class="comment"> * of this software and associated documentation files (the "Software"), to</span></div><div class="line"><a name="l00008"></a><span class="lineno"> 8</span> <span class="comment"> * deal in the Software without restriction, including without limitation the</span></div><div class="line"><a name="l00009"></a><span class="lineno"> 9</span> <span class="comment"> * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or</span></div><div class="line"><a name="l00010"></a><span class="lineno"> 10</span> <span class="comment"> * sell copies of the Software, and to permit persons to whom the Software is</span></div><div class="line"><a name="l00011"></a><span class="lineno"> 11</span> <span class="comment"> * furnished to do so, subject to the following conditions:</span></div><div class="line"><a name="l00012"></a><span class="lineno"> 12</span> <span class="comment"> *</span></div><div class="line"><a name="l00013"></a><span class="lineno"> 13</span> <span class="comment"> * The above copyright notice and this permission notice shall be included in all</span></div><div class="line"><a name="l00014"></a><span class="lineno"> 14</span> <span class="comment"> * copies or substantial portions of the Software.</span></div><div class="line"><a name="l00015"></a><span class="lineno"> 15</span> <span class="comment"> *</span></div><div class="line"><a name="l00016"></a><span class="lineno"> 16</span> <span class="comment"> * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR</span></div><div class="line"><a name="l00017"></a><span class="lineno"> 17</span> <span class="comment"> * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,</span></div><div class="line"><a name="l00018"></a><span class="lineno"> 18</span> <span class="comment"> * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE</span></div><div class="line"><a name="l00019"></a><span class="lineno"> 19</span> <span class="comment"> * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER</span></div><div class="line"><a name="l00020"></a><span class="lineno"> 20</span> <span class="comment"> * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,</span></div><div class="line"><a name="l00021"></a><span class="lineno"> 21</span> <span class="comment"> * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE</span></div><div class="line"><a name="l00022"></a><span class="lineno"> 22</span> <span class="comment"> * SOFTWARE.</span></div><div class="line"><a name="l00023"></a><span class="lineno"> 23</span> <span class="comment"> */</span></div><div class="line"><a name="l00024"></a><span class="lineno"> 24</span> <span class="preprocessor">#ifndef ARM_COMPUTE_CLGEMMLOWPOUTPUTSTAGE_H</span></div><div class="line"><a name="l00025"></a><span class="lineno"> 25</span> <span class="preprocessor">#define ARM_COMPUTE_CLGEMMLOWPOUTPUTSTAGE_H</span></div><div class="line"><a name="l00026"></a><span class="lineno"> 26</span> </div><div class="line"><a name="l00027"></a><span class="lineno"> 27</span> <span class="preprocessor">#include "<a class="code" href="_i_c_l_simple_function_8h.xhtml">arm_compute/runtime/CL/ICLSimpleFunction.h</a>"</span></div><div class="line"><a name="l00028"></a><span class="lineno"> 28</span> <span class="comment"></span></div><div class="line"><a name="l00029"></a><span class="lineno"> 29</span> <span class="comment">/** This file contains all available output stages for GEMMLowp on OpenCL.</span></div><div class="line"><a name="l00030"></a><span class="lineno"> 30</span> <span class="comment"> *</span></div><div class="line"><a name="l00031"></a><span class="lineno"> 31</span> <span class="comment"> * In gemmlowp, the "output stage" is the process that takes a final int32 accumulator value (the output of @ref CLGEMMLowpMatrixMultiplyCore),</span></div><div class="line"><a name="l00032"></a><span class="lineno"> 32</span> <span class="comment"> * and processes it to obtain the final QASYMM8/QASYMM8_SIGNED value.</span></div><div class="line"><a name="l00033"></a><span class="lineno"> 33</span> <span class="comment"> *</span></div><div class="line"><a name="l00034"></a><span class="lineno"> 34</span> <span class="comment"> * More information about the GEMMLowp output stage can be found at https://github.com/google/gemmlowp/blob/master/doc/output.md</span></div><div class="line"><a name="l00035"></a><span class="lineno"> 35</span> <span class="comment"> */</span></div><div class="line"><a name="l00036"></a><span class="lineno"> 36</span> </div><div class="line"><a name="l00037"></a><span class="lineno"> 37</span> <span class="keyword">namespace </span><a class="code" href="namespacearm__compute.xhtml">arm_compute</a></div><div class="line"><a name="l00038"></a><span class="lineno"> 38</span> {</div><div class="line"><a name="l00039"></a><span class="lineno"> 39</span> <span class="keyword">class </span>ITensor;</div><div class="line"><a name="l00040"></a><span class="lineno"> 40</span> <span class="comment"></span></div><div class="line"><a name="l00041"></a><span class="lineno"> 41</span> <span class="comment">/** Basic function to execute CLGEMMLowpQuantizeDownInt32ToUint8Scale on OpenCL.</span></div><div class="line"><a name="l00042"></a><span class="lineno"> 42</span> <span class="comment"> *</span></div><div class="line"><a name="l00043"></a><span class="lineno"> 43</span> <span class="comment"> * CLGEMMLowpQuantizeDownInt32ToUint8Scale depends on 3 parameters: result_offset, result_mult_int, result_shift</span></div><div class="line"><a name="l00044"></a><span class="lineno"> 44</span> <span class="comment"> * The final result is:</span></div><div class="line"><a name="l00045"></a><span class="lineno"> 45</span> <span class="comment"> *</span></div><div class="line"><a name="l00046"></a><span class="lineno"> 46</span> <span class="comment"> * ((input[i][k] + result_offset) * result_mult_int) >> result_shift</span></div><div class="line"><a name="l00047"></a><span class="lineno"> 47</span> <span class="comment"> *</span></div><div class="line"><a name="l00048"></a><span class="lineno"> 48</span> <span class="comment"> * In case the bias tensor is provided, the final result is:</span></div><div class="line"><a name="l00049"></a><span class="lineno"> 49</span> <span class="comment"> *</span></div><div class="line"><a name="l00050"></a><span class="lineno"> 50</span> <span class="comment"> * ((input[i][k] + bias[k] + result_offset) * result_mult_int) >> result_shift</span></div><div class="line"><a name="l00051"></a><span class="lineno"> 51</span> <span class="comment"> *</span></div><div class="line"><a name="l00052"></a><span class="lineno"> 52</span> <span class="comment"> * This function calls the following OpenCL kernels:</span></div><div class="line"><a name="l00053"></a><span class="lineno"> 53</span> <span class="comment"> *</span></div><div class="line"><a name="l00054"></a><span class="lineno"> 54</span> <span class="comment"> * -# @ref CLGEMMLowpQuantizeDownInt32ToUint8ScaleKernel</span></div><div class="line"><a name="l00055"></a><span class="lineno"> 55</span> <span class="comment"> *</span></div><div class="line"><a name="l00056"></a><span class="lineno"> 56</span> <span class="comment"> * @note The function accepts also 2 optional input arguments (min and max) which can be used to implement "rectified linear unit" activation functions</span></div><div class="line"><a name="l00057"></a><span class="lineno"> 57</span> <span class="comment"> * after the result is shifted right by result_shift</span></div><div class="line"><a name="l00058"></a><span class="lineno"> 58</span> <span class="comment">*/</span></div><div class="line"><a name="l00059"></a><span class="lineno"><a class="line" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale.xhtml"> 59</a></span> <span class="keyword">class </span><a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale.xhtml">CLGEMMLowpQuantizeDownInt32ToUint8Scale</a> : <span class="keyword">public</span> <a class="code" href="classarm__compute_1_1_i_c_l_simple_function.xhtml">ICLSimpleFunction</a></div><div class="line"><a name="l00060"></a><span class="lineno"> 60</span> {</div><div class="line"><a name="l00061"></a><span class="lineno"> 61</span> <span class="keyword">public</span>:<span class="comment"></span></div><div class="line"><a name="l00062"></a><span class="lineno"> 62</span> <span class="comment"> /** Initialise the kernel's inputs, output</span></div><div class="line"><a name="l00063"></a><span class="lineno"> 63</span> <span class="comment"> *</span></div><div class="line"><a name="l00064"></a><span class="lineno"> 64</span> <span class="comment"> * @param[in] input Input tensor. It is the output of @ref CLGEMMLowpMatrixMultiplyCore function. Data type supported: S32</span></div><div class="line"><a name="l00065"></a><span class="lineno"> 65</span> <span class="comment"> * @param[in] bias Biases tensor. Only shared biases supported and it can be a nullptr if the addition of biases is not required.</span></div><div class="line"><a name="l00066"></a><span class="lineno"> 66</span> <span class="comment"> * Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p input.</span></div><div class="line"><a name="l00067"></a><span class="lineno"> 67</span> <span class="comment"> * @param[out] output Output tensor. Data type supported: Data type supported: QASYMM8</span></div><div class="line"><a name="l00068"></a><span class="lineno"> 68</span> <span class="comment"> * @param[in] result_offset Offset to be added to each element of the input matrix</span></div><div class="line"><a name="l00069"></a><span class="lineno"> 69</span> <span class="comment"> * @param[in] result_mult_int Value to be multiplied to each element of the input matrix when once the result_offset has been add</span></div><div class="line"><a name="l00070"></a><span class="lineno"> 70</span> <span class="comment"> * @param[in] result_shift Number of bits to shift right the result before converting back to QASYMM8</span></div><div class="line"><a name="l00071"></a><span class="lineno"> 71</span> <span class="comment"> * @param[in] min (Optional) Min value used to saturate down the output result before converting back to QASYMM8</span></div><div class="line"><a name="l00072"></a><span class="lineno"> 72</span> <span class="comment"> * @param[in] max (Optional) Max value used to saturate up the output result before converting back to QASYMM8,</span></div><div class="line"><a name="l00073"></a><span class="lineno"> 73</span> <span class="comment"> * Along with @p min, this value can be used to implement "rectified linear unit" activation functions</span></div><div class="line"><a name="l00074"></a><span class="lineno"> 74</span> <span class="comment"> */</span></div><div class="line"><a name="l00075"></a><span class="lineno"> 75</span>  <span class="keywordtype">void</span> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale.xhtml#af127cda81674a7864000a3e70877c4c2">configure</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a8fcf2ddd9a1d58b1b280f5c0aed71845">input</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3a77be8aebd8e00522b32061d46ccdbd">bias</a>, <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *output, <span class="keywordtype">int</span> result_offset, <span class="keywordtype">int</span> result_mult_int, <span class="keywordtype">int</span> result_shift, <span class="keywordtype">int</span> min = 0, <span class="keywordtype">int</span> max = 0);<span class="comment"></span></div><div class="line"><a name="l00076"></a><span class="lineno"> 76</span> <span class="comment"> /** Static function to check if given info will lead to a valid configuration of @ref CLGEMMLowpQuantizeDownInt32ToUint8Scale</span></div><div class="line"><a name="l00077"></a><span class="lineno"> 77</span> <span class="comment"> *</span></div><div class="line"><a name="l00078"></a><span class="lineno"> 78</span> <span class="comment"> * @param[in] input Input tensor. It is the output of @ref CLGEMMLowpMatrixMultiplyCore function. Data type supported: S32</span></div><div class="line"><a name="l00079"></a><span class="lineno"> 79</span> <span class="comment"> * @param[in] bias Biases tensor. Only shared biases supported and it can be a nullptr if the addition of biases is not required.</span></div><div class="line"><a name="l00080"></a><span class="lineno"> 80</span> <span class="comment"> * Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p input.</span></div><div class="line"><a name="l00081"></a><span class="lineno"> 81</span> <span class="comment"> * @param[in] output Output tensor. Data type supported: Data type supported: QASYMM8</span></div><div class="line"><a name="l00082"></a><span class="lineno"> 82</span> <span class="comment"> * @param[in] min (Optional) Min value used to saturate down the output result before converting back to QASYMM8</span></div><div class="line"><a name="l00083"></a><span class="lineno"> 83</span> <span class="comment"> * @param[in] max (Optional) Max value used to saturate up the output result before converting back to QASYMM8,</span></div><div class="line"><a name="l00084"></a><span class="lineno"> 84</span> <span class="comment"> * Along with @p min, this value can be used to implement "rectified linear unit" activation functions</span></div><div class="line"><a name="l00085"></a><span class="lineno"> 85</span> <span class="comment"> *</span></div><div class="line"><a name="l00086"></a><span class="lineno"> 86</span> <span class="comment"> * @return a status</span></div><div class="line"><a name="l00087"></a><span class="lineno"> 87</span> <span class="comment"> */</span></div><div class="line"><a name="l00088"></a><span class="lineno"> 88</span>  <span class="keyword">static</span> <a class="code" href="classarm__compute_1_1_status.xhtml">Status</a> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale.xhtml#aee63e7671cf04d15be2da1b83d90e61b">validate</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a8fcf2ddd9a1d58b1b280f5c0aed71845">input</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3a77be8aebd8e00522b32061d46ccdbd">bias</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *output, <span class="keywordtype">int</span> min = 0, <span class="keywordtype">int</span> max = 0);</div><div class="line"><a name="l00089"></a><span class="lineno"> 89</span> };</div><div class="line"><a name="l00090"></a><span class="lineno"> 90</span> <span class="comment"></span></div><div class="line"><a name="l00091"></a><span class="lineno"> 91</span> <span class="comment">/** Basic function to execute CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint on OpenCL.</span></div><div class="line"><a name="l00092"></a><span class="lineno"> 92</span> <span class="comment"> *</span></div><div class="line"><a name="l00093"></a><span class="lineno"> 93</span> <span class="comment"> * CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint depends on 3 parameters:</span></div><div class="line"><a name="l00094"></a><span class="lineno"> 94</span> <span class="comment"> *</span></div><div class="line"><a name="l00095"></a><span class="lineno"> 95</span> <span class="comment"> * result_fixedpoint_multiplier, result_shift, result_offset_after_shift</span></div><div class="line"><a name="l00096"></a><span class="lineno"> 96</span> <span class="comment"> *</span></div><div class="line"><a name="l00097"></a><span class="lineno"> 97</span> <span class="comment"> * The final result is:</span></div><div class="line"><a name="l00098"></a><span class="lineno"> 98</span> <span class="comment"> *</span></div><div class="line"><a name="l00099"></a><span class="lineno"> 99</span> <span class="comment"> * (FixedPointMul(input[i][k], result_fixedpoint_multiplier) >> result_shift) + result_offset_after_shift</span></div><div class="line"><a name="l00100"></a><span class="lineno"> 100</span> <span class="comment"> *</span></div><div class="line"><a name="l00101"></a><span class="lineno"> 101</span> <span class="comment"> * where FixedPointMul(x, y) is the nearest integer to the following</span></div><div class="line"><a name="l00102"></a><span class="lineno"> 102</span> <span class="comment"> * mathematical expression, evaluated without overflow or intermediate rounding:</span></div><div class="line"><a name="l00103"></a><span class="lineno"> 103</span> <span class="comment"> *</span></div><div class="line"><a name="l00104"></a><span class="lineno"> 104</span> <span class="comment"> * (x * y) / 2^31</span></div><div class="line"><a name="l00105"></a><span class="lineno"> 105</span> <span class="comment"> *</span></div><div class="line"><a name="l00106"></a><span class="lineno"> 106</span> <span class="comment"> * For more information: https://github.com/google/gemmlowp/blob/master/public/output_stages.h#L68</span></div><div class="line"><a name="l00107"></a><span class="lineno"> 107</span> <span class="comment"> *</span></div><div class="line"><a name="l00108"></a><span class="lineno"> 108</span> <span class="comment"> * In case the bias tensor is provided, the final result is:</span></div><div class="line"><a name="l00109"></a><span class="lineno"> 109</span> <span class="comment"> *</span></div><div class="line"><a name="l00110"></a><span class="lineno"> 110</span> <span class="comment"> * ((FixedPointMul(input[i][k] + bias[k], result_fixedpoint_multiplier)) >> result_shift) + result_offset_after_shift</span></div><div class="line"><a name="l00111"></a><span class="lineno"> 111</span> <span class="comment"> *</span></div><div class="line"><a name="l00112"></a><span class="lineno"> 112</span> <span class="comment"> * This function calls the following OpenCL kernels:</span></div><div class="line"><a name="l00113"></a><span class="lineno"> 113</span> <span class="comment"> *</span></div><div class="line"><a name="l00114"></a><span class="lineno"> 114</span> <span class="comment"> * -# @ref CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPointKernel</span></div><div class="line"><a name="l00115"></a><span class="lineno"> 115</span> <span class="comment"> *</span></div><div class="line"><a name="l00116"></a><span class="lineno"> 116</span> <span class="comment"> * @note The function accepts also 2 optional input arguments (min and max) which can be used to implement "rectified linear unit" activation functions</span></div><div class="line"><a name="l00117"></a><span class="lineno"> 117</span> <span class="comment"> * after the result is shifted right by result_shift</span></div><div class="line"><a name="l00118"></a><span class="lineno"> 118</span> <span class="comment">*/</span></div><div class="line"><a name="l00119"></a><span class="lineno"><a class="line" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_fixed_point.xhtml"> 119</a></span> <span class="keyword">class </span><a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_fixed_point.xhtml">CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint</a> : <span class="keyword">public</span> <a class="code" href="classarm__compute_1_1_i_c_l_simple_function.xhtml">ICLSimpleFunction</a></div><div class="line"><a name="l00120"></a><span class="lineno"> 120</span> {</div><div class="line"><a name="l00121"></a><span class="lineno"> 121</span> <span class="keyword">public</span>:<span class="comment"></span></div><div class="line"><a name="l00122"></a><span class="lineno"> 122</span> <span class="comment"> /** Initialise the kernel's inputs, output</span></div><div class="line"><a name="l00123"></a><span class="lineno"> 123</span> <span class="comment"> *</span></div><div class="line"><a name="l00124"></a><span class="lineno"> 124</span> <span class="comment"> * @param[in] input Input tensor. Data type supported: S32</span></div><div class="line"><a name="l00125"></a><span class="lineno"> 125</span> <span class="comment"> * @param[in] bias Biases tensor. Only shared biases supported and it can be a nullptr if the biases addition is not required.</span></div><div class="line"><a name="l00126"></a><span class="lineno"> 126</span> <span class="comment"> * Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p input.</span></div><div class="line"><a name="l00127"></a><span class="lineno"> 127</span> <span class="comment"> * @param[out] output Output tensor. Data type supported: Data type supported: QASYMM8</span></div><div class="line"><a name="l00128"></a><span class="lineno"> 128</span> <span class="comment"> * @param[in] result_fixedpoint_multiplier Fixed point value to be multiplied to each element of the input matrix when once the result_offset has been add</span></div><div class="line"><a name="l00129"></a><span class="lineno"> 129</span> <span class="comment"> * @param[in] result_shift Number of bits to shift right the result after the fixed point multiplication</span></div><div class="line"><a name="l00130"></a><span class="lineno"> 130</span> <span class="comment"> * @param[in] result_offset_after_shift Offset to be applied to result before converting it back to QASYMM8</span></div><div class="line"><a name="l00131"></a><span class="lineno"> 131</span> <span class="comment"> * @param[in] min (Optional) Min value used to saturate down the output result before converting back to QASYMM8</span></div><div class="line"><a name="l00132"></a><span class="lineno"> 132</span> <span class="comment"> * @param[in] max (Optional) Max value used to saturate up the output result before converting back to QASYMM8,</span></div><div class="line"><a name="l00133"></a><span class="lineno"> 133</span> <span class="comment"> * Along with @p min, this value can be used to implement "rectified linear unit" activation functions</span></div><div class="line"><a name="l00134"></a><span class="lineno"> 134</span> <span class="comment"> */</span></div><div class="line"><a name="l00135"></a><span class="lineno"> 135</span>  <span class="keywordtype">void</span> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_fixed_point.xhtml#a15da37f661fdcd81c1f25c5d6bdc6abd">configure</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a8fcf2ddd9a1d58b1b280f5c0aed71845">input</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3a77be8aebd8e00522b32061d46ccdbd">bias</a>, <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *output, <span class="keywordtype">int</span> result_fixedpoint_multiplier, <span class="keywordtype">int</span> result_shift, <span class="keywordtype">int</span> result_offset_after_shift,</div><div class="line"><a name="l00136"></a><span class="lineno"> 136</span>  <span class="keywordtype">int</span> min = 0, <span class="keywordtype">int</span> max = 0);<span class="comment"></span></div><div class="line"><a name="l00137"></a><span class="lineno"> 137</span> <span class="comment"> /** Static function to check if given info will lead to a valid configuration of @ref CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint</span></div><div class="line"><a name="l00138"></a><span class="lineno"> 138</span> <span class="comment"> *</span></div><div class="line"><a name="l00139"></a><span class="lineno"> 139</span> <span class="comment"> * @param[in] input Input tensor. It is the output of @ref CLGEMMLowpMatrixMultiplyCore function. Data type supported: S32</span></div><div class="line"><a name="l00140"></a><span class="lineno"> 140</span> <span class="comment"> * @param[in] bias Biases tensor. Only shared biases supported and it can be a nullptr if the addition of biases is not required.</span></div><div class="line"><a name="l00141"></a><span class="lineno"> 141</span> <span class="comment"> * Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p input.</span></div><div class="line"><a name="l00142"></a><span class="lineno"> 142</span> <span class="comment"> * @param[in] output Output tensor. Data type supported: Data type supported: QASYMM8</span></div><div class="line"><a name="l00143"></a><span class="lineno"> 143</span> <span class="comment"> * @param[in] min (Optional) Min value used to saturate down the output result before converting back to QASYMM8</span></div><div class="line"><a name="l00144"></a><span class="lineno"> 144</span> <span class="comment"> * @param[in] max (Optional) Max value used to saturate up the output result before converting back to QASYMM8,</span></div><div class="line"><a name="l00145"></a><span class="lineno"> 145</span> <span class="comment"> * Along with @p min, this value can be used to implement "rectified linear unit" activation functions</span></div><div class="line"><a name="l00146"></a><span class="lineno"> 146</span> <span class="comment"> *</span></div><div class="line"><a name="l00147"></a><span class="lineno"> 147</span> <span class="comment"> * @return a status</span></div><div class="line"><a name="l00148"></a><span class="lineno"> 148</span> <span class="comment"> */</span></div><div class="line"><a name="l00149"></a><span class="lineno"> 149</span>  <span class="keyword">static</span> <a class="code" href="classarm__compute_1_1_status.xhtml">Status</a> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_fixed_point.xhtml#aee63e7671cf04d15be2da1b83d90e61b">validate</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a8fcf2ddd9a1d58b1b280f5c0aed71845">input</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3a77be8aebd8e00522b32061d46ccdbd">bias</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *output, <span class="keywordtype">int</span> min = 0, <span class="keywordtype">int</span> max = 0);</div><div class="line"><a name="l00150"></a><span class="lineno"> 150</span> };</div><div class="line"><a name="l00151"></a><span class="lineno"> 151</span> <span class="comment"></span></div><div class="line"><a name="l00152"></a><span class="lineno"> 152</span> <span class="comment">/** Basic function to execute CLGEMMLowpQuantizeDownInt32ToInt8ScaleByFixedPoint on OpenCL.</span></div><div class="line"><a name="l00153"></a><span class="lineno"> 153</span> <span class="comment"> *</span></div><div class="line"><a name="l00154"></a><span class="lineno"> 154</span> <span class="comment"> * CLGEMMLowpQuantizeDownInt32ToInt8ScaleByFixedPoint depends on 3 parameters:</span></div><div class="line"><a name="l00155"></a><span class="lineno"> 155</span> <span class="comment"> *</span></div><div class="line"><a name="l00156"></a><span class="lineno"> 156</span> <span class="comment"> * result_fixedpoint_multiplier, result_shift, result_offset_after_shift</span></div><div class="line"><a name="l00157"></a><span class="lineno"> 157</span> <span class="comment"> *</span></div><div class="line"><a name="l00158"></a><span class="lineno"> 158</span> <span class="comment"> * The final result is:</span></div><div class="line"><a name="l00159"></a><span class="lineno"> 159</span> <span class="comment"> *</span></div><div class="line"><a name="l00160"></a><span class="lineno"> 160</span> <span class="comment"> * (FixedPointMul(input[i][k], result_fixedpoint_multiplier) >> result_shift) + result_offset_after_shift</span></div><div class="line"><a name="l00161"></a><span class="lineno"> 161</span> <span class="comment"> *</span></div><div class="line"><a name="l00162"></a><span class="lineno"> 162</span> <span class="comment"> * where FixedPointMul(x, y) is the nearest integer to the following</span></div><div class="line"><a name="l00163"></a><span class="lineno"> 163</span> <span class="comment"> * mathematical expression, evaluated without overflow or intermediate rounding:</span></div><div class="line"><a name="l00164"></a><span class="lineno"> 164</span> <span class="comment"> *</span></div><div class="line"><a name="l00165"></a><span class="lineno"> 165</span> <span class="comment"> * (x * y) / 2^31</span></div><div class="line"><a name="l00166"></a><span class="lineno"> 166</span> <span class="comment"> *</span></div><div class="line"><a name="l00167"></a><span class="lineno"> 167</span> <span class="comment"> * For more information: https://github.com/google/gemmlowp/blob/master/public/output_stages.h#L68</span></div><div class="line"><a name="l00168"></a><span class="lineno"> 168</span> <span class="comment"> *</span></div><div class="line"><a name="l00169"></a><span class="lineno"> 169</span> <span class="comment"> * In case the bias tensor is provided, the final result is:</span></div><div class="line"><a name="l00170"></a><span class="lineno"> 170</span> <span class="comment"> *</span></div><div class="line"><a name="l00171"></a><span class="lineno"> 171</span> <span class="comment"> * ((FixedPointMul(input[i][k] + bias[k], result_fixedpoint_multiplier)) >> result_shift) + result_offset_after_shift</span></div><div class="line"><a name="l00172"></a><span class="lineno"> 172</span> <span class="comment"> *</span></div><div class="line"><a name="l00173"></a><span class="lineno"> 173</span> <span class="comment"> * This function calls the following OpenCL kernels:</span></div><div class="line"><a name="l00174"></a><span class="lineno"> 174</span> <span class="comment"> *</span></div><div class="line"><a name="l00175"></a><span class="lineno"> 175</span> <span class="comment"> * -# @ref CLGEMMLowpQuantizeDownInt32ToInt8ScaleByFixedPointKernel</span></div><div class="line"><a name="l00176"></a><span class="lineno"> 176</span> <span class="comment"> *</span></div><div class="line"><a name="l00177"></a><span class="lineno"> 177</span> <span class="comment"> * @note The function accepts also 2 optional input arguments (min and max) which can be used to implement "rectified linear unit" activation functions</span></div><div class="line"><a name="l00178"></a><span class="lineno"> 178</span> <span class="comment"> * after the result is shifted right by result_shift</span></div><div class="line"><a name="l00179"></a><span class="lineno"> 179</span> <span class="comment">*/</span></div><div class="line"><a name="l00180"></a><span class="lineno"><a class="line" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int8_scale_by_fixed_point.xhtml"> 180</a></span> <span class="keyword">class </span><a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int8_scale_by_fixed_point.xhtml">CLGEMMLowpQuantizeDownInt32ToInt8ScaleByFixedPoint</a> : <span class="keyword">public</span> <a class="code" href="classarm__compute_1_1_i_c_l_simple_function.xhtml">ICLSimpleFunction</a></div><div class="line"><a name="l00181"></a><span class="lineno"> 181</span> {</div><div class="line"><a name="l00182"></a><span class="lineno"> 182</span> <span class="keyword">public</span>:<span class="comment"></span></div><div class="line"><a name="l00183"></a><span class="lineno"> 183</span> <span class="comment"> /** Initialise the kernel's inputs, output</span></div><div class="line"><a name="l00184"></a><span class="lineno"> 184</span> <span class="comment"> *</span></div><div class="line"><a name="l00185"></a><span class="lineno"> 185</span> <span class="comment"> * @param[in] input Input tensor. Data type supported: S32</span></div><div class="line"><a name="l00186"></a><span class="lineno"> 186</span> <span class="comment"> * @param[in] bias Biases tensor. Only shared biases supported and it can be a nullptr if the biases addition is not required.</span></div><div class="line"><a name="l00187"></a><span class="lineno"> 187</span> <span class="comment"> * Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p input.</span></div><div class="line"><a name="l00188"></a><span class="lineno"> 188</span> <span class="comment"> * @param[out] output Output tensor. Data type supported: Data type supported: QASYMM8_SIGNED</span></div><div class="line"><a name="l00189"></a><span class="lineno"> 189</span> <span class="comment"> * @param[in] result_fixedpoint_multiplier Fixed point value to be multiplied to each element of the input matrix when once the result_offset has been add</span></div><div class="line"><a name="l00190"></a><span class="lineno"> 190</span> <span class="comment"> * @param[in] result_shift Number of bits to shift right the result after the fixed point multiplication</span></div><div class="line"><a name="l00191"></a><span class="lineno"> 191</span> <span class="comment"> * @param[in] result_offset_after_shift Offset to be applied to result before converting it back to QASYMM8_SIGNED</span></div><div class="line"><a name="l00192"></a><span class="lineno"> 192</span> <span class="comment"> * @param[in] min (Optional) Min value used to saturate down the output result before converting back to QASYMM8_SIGNED. Defaults to 0</span></div><div class="line"><a name="l00193"></a><span class="lineno"> 193</span> <span class="comment"> * @param[in] max (Optional) Max value used to saturate up the output result before converting back to QASYMM8_SIGNED. Defaults to 0</span></div><div class="line"><a name="l00194"></a><span class="lineno"> 194</span> <span class="comment"> * Along with @p min, this value can be used to implement "rectified linear unit" activation functions</span></div><div class="line"><a name="l00195"></a><span class="lineno"> 195</span> <span class="comment"> */</span></div><div class="line"><a name="l00196"></a><span class="lineno"> 196</span>  <span class="keywordtype">void</span> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int8_scale_by_fixed_point.xhtml#a15da37f661fdcd81c1f25c5d6bdc6abd">configure</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a8fcf2ddd9a1d58b1b280f5c0aed71845">input</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3a77be8aebd8e00522b32061d46ccdbd">bias</a>, <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *output, <span class="keywordtype">int</span> result_fixedpoint_multiplier, <span class="keywordtype">int</span> result_shift, <span class="keywordtype">int</span> result_offset_after_shift,</div><div class="line"><a name="l00197"></a><span class="lineno"> 197</span>  <span class="keywordtype">int</span> min = 0, <span class="keywordtype">int</span> max = 0);<span class="comment"></span></div><div class="line"><a name="l00198"></a><span class="lineno"> 198</span> <span class="comment"> /** Static function to check if given info will lead to a valid configuration of @ref CLGEMMLowpQuantizeDownInt32ToInt8ScaleByFixedPoint</span></div><div class="line"><a name="l00199"></a><span class="lineno"> 199</span> <span class="comment"> *</span></div><div class="line"><a name="l00200"></a><span class="lineno"> 200</span> <span class="comment"> * @param[in] input Input tensor. It is the output of @ref CLGEMMLowpMatrixMultiplyCore function. Data type supported: S32</span></div><div class="line"><a name="l00201"></a><span class="lineno"> 201</span> <span class="comment"> * @param[in] bias Biases tensor. Only shared biases supported and it can be a nullptr if the addition of biases is not required.</span></div><div class="line"><a name="l00202"></a><span class="lineno"> 202</span> <span class="comment"> * Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p input.</span></div><div class="line"><a name="l00203"></a><span class="lineno"> 203</span> <span class="comment"> * @param[in] output Output tensor. Data type supported: Data type supported: QASYMM8_SIGNED</span></div><div class="line"><a name="l00204"></a><span class="lineno"> 204</span> <span class="comment"> * @param[in] min (Optional) Min value used to saturate down the output result before converting back to QASYMM8_SIGNED. Defaults to 0</span></div><div class="line"><a name="l00205"></a><span class="lineno"> 205</span> <span class="comment"> * @param[in] max (Optional) Max value used to saturate up the output result before converting back to QASYMM8_SIGNED. Defaults to 0</span></div><div class="line"><a name="l00206"></a><span class="lineno"> 206</span> <span class="comment"> * Along with @p min, this value can be used to implement "rectified linear unit" activation functions</span></div><div class="line"><a name="l00207"></a><span class="lineno"> 207</span> <span class="comment"> *</span></div><div class="line"><a name="l00208"></a><span class="lineno"> 208</span> <span class="comment"> * @return a status</span></div><div class="line"><a name="l00209"></a><span class="lineno"> 209</span> <span class="comment"> */</span></div><div class="line"><a name="l00210"></a><span class="lineno"> 210</span>  <span class="keyword">static</span> <a class="code" href="classarm__compute_1_1_status.xhtml">Status</a> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int8_scale_by_fixed_point.xhtml#aee63e7671cf04d15be2da1b83d90e61b">validate</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a8fcf2ddd9a1d58b1b280f5c0aed71845">input</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3a77be8aebd8e00522b32061d46ccdbd">bias</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *output, <span class="keywordtype">int</span> min = 0, <span class="keywordtype">int</span> max = 0);</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"></span></div><div class="line"><a name="l00213"></a><span class="lineno"> 213</span> <span class="comment">/** Basic function to execute CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloat on OpenCL.</span></div><div class="line"><a name="l00214"></a><span class="lineno"> 214</span> <span class="comment"> *</span></div><div class="line"><a name="l00215"></a><span class="lineno"> 215</span> <span class="comment"> * This function calls the following OpenCL kernels:</span></div><div class="line"><a name="l00216"></a><span class="lineno"> 216</span> <span class="comment"> *</span></div><div class="line"><a name="l00217"></a><span class="lineno"> 217</span> <span class="comment"> * -# @ref CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloatKernel</span></div><div class="line"><a name="l00218"></a><span class="lineno"> 218</span> <span class="comment"> *</span></div><div class="line"><a name="l00219"></a><span class="lineno"> 219</span> <span class="comment"> * @note The function accepts also 2 optional input arguments (min and max) which can be used to implement "rectified linear unit" activation functions</span></div><div class="line"><a name="l00220"></a><span class="lineno"> 220</span> <span class="comment"> * after the result is shifted right by result_shift</span></div><div class="line"><a name="l00221"></a><span class="lineno"> 221</span> <span class="comment">*/</span></div><div class="line"><a name="l00222"></a><span class="lineno"><a class="line" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_float.xhtml"> 222</a></span> <span class="keyword">class </span><a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_float.xhtml">CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloat</a> : <span class="keyword">public</span> <a class="code" href="classarm__compute_1_1_i_c_l_simple_function.xhtml">ICLSimpleFunction</a></div><div class="line"><a name="l00223"></a><span class="lineno"> 223</span> {</div><div class="line"><a name="l00224"></a><span class="lineno"> 224</span> <span class="keyword">public</span>:<span class="comment"></span></div><div class="line"><a name="l00225"></a><span class="lineno"> 225</span> <span class="comment"> /** Initialise the kernel's inputs, output</span></div><div class="line"><a name="l00226"></a><span class="lineno"> 226</span> <span class="comment"> *</span></div><div class="line"><a name="l00227"></a><span class="lineno"> 227</span> <span class="comment"> * @param[in] input Input tensor. Data type supported: S32</span></div><div class="line"><a name="l00228"></a><span class="lineno"> 228</span> <span class="comment"> * @param[in] bias Biases tensor. Only shared biases supported and it can be a nullptr if the biases addition is not required.</span></div><div class="line"><a name="l00229"></a><span class="lineno"> 229</span> <span class="comment"> * Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p input.</span></div><div class="line"><a name="l00230"></a><span class="lineno"> 230</span> <span class="comment"> * @param[out] output Output tensor. Data type supported: Data type supported: QASYMM8</span></div><div class="line"><a name="l00231"></a><span class="lineno"> 231</span> <span class="comment"> * @param[in] multiplier Float multiplier to be multiplied to each element of the input matrix</span></div><div class="line"><a name="l00232"></a><span class="lineno"> 232</span> <span class="comment"> * @param[in] offset Offset to be applied to result before converting it back to QASYMM8</span></div><div class="line"><a name="l00233"></a><span class="lineno"> 233</span> <span class="comment"> * @param[in] min (Optional) Min value used to saturate down the output result before converting back to QASYMM8</span></div><div class="line"><a name="l00234"></a><span class="lineno"> 234</span> <span class="comment"> * @param[in] max (Optional) Max value used to saturate up the output result before converting back to QASYMM8,</span></div><div class="line"><a name="l00235"></a><span class="lineno"> 235</span> <span class="comment"> * Along with @p min, this value can be used to implement "rectified linear unit" activation functions</span></div><div class="line"><a name="l00236"></a><span class="lineno"> 236</span> <span class="comment"> */</span></div><div class="line"><a name="l00237"></a><span class="lineno"> 237</span>  <span class="keywordtype">void</span> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_float.xhtml#a49bf33836c645a5b5aa385411ab56964">configure</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a8fcf2ddd9a1d58b1b280f5c0aed71845">input</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3a77be8aebd8e00522b32061d46ccdbd">bias</a>, <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *output, <span class="keywordtype">float</span> multiplier, <span class="keywordtype">int</span> <a class="code" href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a009469e4d9b8fce3b6d5e97d2077827d">offset</a>, <span class="keywordtype">int</span> min = 0, <span class="keywordtype">int</span> max = 0);<span class="comment"></span></div><div class="line"><a name="l00238"></a><span class="lineno"> 238</span> <span class="comment"> /** Static function to check if given info will lead to a valid configuration of @ref CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint</span></div><div class="line"><a name="l00239"></a><span class="lineno"> 239</span> <span class="comment"> *</span></div><div class="line"><a name="l00240"></a><span class="lineno"> 240</span> <span class="comment"> * @param[in] input Input tensor. It is the output of @ref CLGEMMLowpMatrixMultiplyCore function. Data type supported: S32</span></div><div class="line"><a name="l00241"></a><span class="lineno"> 241</span> <span class="comment"> * @param[in] bias Biases tensor. Only shared biases supported and it can be a nullptr if the addition of biases is not required.</span></div><div class="line"><a name="l00242"></a><span class="lineno"> 242</span> <span class="comment"> * Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p input.</span></div><div class="line"><a name="l00243"></a><span class="lineno"> 243</span> <span class="comment"> * @param[in] output Output tensor. Data type supported: Data type supported: QASYMM8</span></div><div class="line"><a name="l00244"></a><span class="lineno"> 244</span> <span class="comment"> * @param[in] min (Optional) Min value used to saturate down the output result before converting back to QASYMM8</span></div><div class="line"><a name="l00245"></a><span class="lineno"> 245</span> <span class="comment"> * @param[in] max (Optional) Max value used to saturate up the output result before converting back to QASYMM8,</span></div><div class="line"><a name="l00246"></a><span class="lineno"> 246</span> <span class="comment"> * Along with @p min, this value can be used to implement "rectified linear unit" activation functions</span></div><div class="line"><a name="l00247"></a><span class="lineno"> 247</span> <span class="comment"> *</span></div><div class="line"><a name="l00248"></a><span class="lineno"> 248</span> <span class="comment"> * @return a status</span></div><div class="line"><a name="l00249"></a><span class="lineno"> 249</span> <span class="comment"> */</span></div><div class="line"><a name="l00250"></a><span class="lineno"> 250</span>  <span class="keyword">static</span> <a class="code" href="classarm__compute_1_1_status.xhtml">Status</a> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_float.xhtml#aee63e7671cf04d15be2da1b83d90e61b">validate</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a8fcf2ddd9a1d58b1b280f5c0aed71845">input</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3a77be8aebd8e00522b32061d46ccdbd">bias</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *output, <span class="keywordtype">int</span> min = 0, <span class="keywordtype">int</span> max = 0);</div><div class="line"><a name="l00251"></a><span class="lineno"> 251</span> };<span class="comment"></span></div><div class="line"><a name="l00252"></a><span class="lineno"> 252</span> <span class="comment">/** Basic function to execute CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPoint on OpenCL.</span></div><div class="line"><a name="l00253"></a><span class="lineno"> 253</span> <span class="comment"> *</span></div><div class="line"><a name="l00254"></a><span class="lineno"> 254</span> <span class="comment"> * CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPoint depends on 2 parameters:</span></div><div class="line"><a name="l00255"></a><span class="lineno"> 255</span> <span class="comment"> *</span></div><div class="line"><a name="l00256"></a><span class="lineno"> 256</span> <span class="comment"> * result_fixedpoint_multiplier, result_shift</span></div><div class="line"><a name="l00257"></a><span class="lineno"> 257</span> <span class="comment"> *</span></div><div class="line"><a name="l00258"></a><span class="lineno"> 258</span> <span class="comment"> * The final result is:</span></div><div class="line"><a name="l00259"></a><span class="lineno"> 259</span> <span class="comment"> *</span></div><div class="line"><a name="l00260"></a><span class="lineno"> 260</span> <span class="comment"> * (FixedPointMul(input[i][k], result_fixedpoint_multiplier) >> result_shift)</span></div><div class="line"><a name="l00261"></a><span class="lineno"> 261</span> <span class="comment"> *</span></div><div class="line"><a name="l00262"></a><span class="lineno"> 262</span> <span class="comment"> * where FixedPointMul(x, y) is the nearest integer to the following</span></div><div class="line"><a name="l00263"></a><span class="lineno"> 263</span> <span class="comment"> * mathematical expression, evaluated without overflow or intermediate rounding:</span></div><div class="line"><a name="l00264"></a><span class="lineno"> 264</span> <span class="comment"> *</span></div><div class="line"><a name="l00265"></a><span class="lineno"> 265</span> <span class="comment"> * (x * y) / 2^31</span></div><div class="line"><a name="l00266"></a><span class="lineno"> 266</span> <span class="comment"> *</span></div><div class="line"><a name="l00267"></a><span class="lineno"> 267</span> <span class="comment"> * For more information: https://github.com/google/gemmlowp/blob/master/public/output_stages.h#L68</span></div><div class="line"><a name="l00268"></a><span class="lineno"> 268</span> <span class="comment"> *</span></div><div class="line"><a name="l00269"></a><span class="lineno"> 269</span> <span class="comment"> * In case the bias tensor is provided, the final result is:</span></div><div class="line"><a name="l00270"></a><span class="lineno"> 270</span> <span class="comment"> *</span></div><div class="line"><a name="l00271"></a><span class="lineno"> 271</span> <span class="comment"> * ((FixedPointMul(input[i][k] + bias[k], result_fixedpoint_multiplier)) >> result_shift) + result_offset_after_shift</span></div><div class="line"><a name="l00272"></a><span class="lineno"> 272</span> <span class="comment"> *</span></div><div class="line"><a name="l00273"></a><span class="lineno"> 273</span> <span class="comment"> * This function calls the following NEON kernels:</span></div><div class="line"><a name="l00274"></a><span class="lineno"> 274</span> <span class="comment"> *</span></div><div class="line"><a name="l00275"></a><span class="lineno"> 275</span> <span class="comment"> * -# @ref CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPointKernel</span></div><div class="line"><a name="l00276"></a><span class="lineno"> 276</span> <span class="comment"> *</span></div><div class="line"><a name="l00277"></a><span class="lineno"> 277</span> <span class="comment"> * @note The function accepts also 2 optional input arguments (min and max) which can be used to implement "rectified linear unit" activation functions</span></div><div class="line"><a name="l00278"></a><span class="lineno"> 278</span> <span class="comment"> * after the result is shifted right by result_shift</span></div><div class="line"><a name="l00279"></a><span class="lineno"> 279</span> <span class="comment">*/</span></div><div class="line"><a name="l00280"></a><span class="lineno"><a class="line" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int16_scale_by_fixed_point.xhtml"> 280</a></span> <span class="keyword">class </span><a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int16_scale_by_fixed_point.xhtml">CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPoint</a> : <span class="keyword">public</span> <a class="code" href="classarm__compute_1_1_i_c_l_simple_function.xhtml">ICLSimpleFunction</a></div><div class="line"><a name="l00281"></a><span class="lineno"> 281</span> {</div><div class="line"><a name="l00282"></a><span class="lineno"> 282</span> <span class="keyword">public</span>:<span class="comment"></span></div><div class="line"><a name="l00283"></a><span class="lineno"> 283</span> <span class="comment"> /** Initialise the kernel's inputs, output</span></div><div class="line"><a name="l00284"></a><span class="lineno"> 284</span> <span class="comment"> *</span></div><div class="line"><a name="l00285"></a><span class="lineno"> 285</span> <span class="comment"> * @param[in] input Input tensor. Data type supported: S32</span></div><div class="line"><a name="l00286"></a><span class="lineno"> 286</span> <span class="comment"> * @param[in] bias Biases tensor. Only shared biases supported and it can be a nullptr if the biases addition is not required.</span></div><div class="line"><a name="l00287"></a><span class="lineno"> 287</span> <span class="comment"> * Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p input.</span></div><div class="line"><a name="l00288"></a><span class="lineno"> 288</span> <span class="comment"> * @param[out] output Output tensor. Data type supported: Data type supported: QSYMM16</span></div><div class="line"><a name="l00289"></a><span class="lineno"> 289</span> <span class="comment"> * @param[in] result_fixedpoint_multiplier Fixed point value to be multiplied to each element of the input matrix when once the result_offset has been add</span></div><div class="line"><a name="l00290"></a><span class="lineno"> 290</span> <span class="comment"> * @param[in] result_shift Number of bits to shift right the result after the fixed point multiplication</span></div><div class="line"><a name="l00291"></a><span class="lineno"> 291</span> <span class="comment"> * @param[in] min (Optional) Min value used to saturate down the output result before converting back to QSYMM16. Defaults to 0.</span></div><div class="line"><a name="l00292"></a><span class="lineno"> 292</span> <span class="comment"> * @param[in] max (Optional) Max value used to saturate up the output result before converting back to QSYMM16.</span></div><div class="line"><a name="l00293"></a><span class="lineno"> 293</span> <span class="comment"> * Along with @p min, this value can be used to implement "rectified linear unit" activation functions. Defaults to 0.</span></div><div class="line"><a name="l00294"></a><span class="lineno"> 294</span> <span class="comment"> */</span></div><div class="line"><a name="l00295"></a><span class="lineno"> 295</span>  <span class="keywordtype">void</span> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int16_scale_by_fixed_point.xhtml#a5c7abeeecc0f2d6c50cc1dafaadf7c19">configure</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a8fcf2ddd9a1d58b1b280f5c0aed71845">input</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3a77be8aebd8e00522b32061d46ccdbd">bias</a>, <a class="code" href="classarm__compute_1_1_i_c_l_tensor.xhtml">ICLTensor</a> *output, <span class="keywordtype">int</span> result_fixedpoint_multiplier, <span class="keywordtype">int</span> result_shift, <span class="keywordtype">int</span> min = 0, <span class="keywordtype">int</span> max = 0);<span class="comment"></span></div><div class="line"><a name="l00296"></a><span class="lineno"> 296</span> <span class="comment"> /** Static function to check if given info will lead to a valid configuration of @ref CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPoint</span></div><div class="line"><a name="l00297"></a><span class="lineno"> 297</span> <span class="comment"> *</span></div><div class="line"><a name="l00298"></a><span class="lineno"> 298</span> <span class="comment"> * @param[in] input Input tensor info. It is the output of @ref CLGEMMLowpMatrixMultiplyCore function. Data type supported: S32</span></div><div class="line"><a name="l00299"></a><span class="lineno"> 299</span> <span class="comment"> * @param[in] bias Biases tensor info. Only shared biases supported and it can be a nullptr if the addition of biases is not required.</span></div><div class="line"><a name="l00300"></a><span class="lineno"> 300</span> <span class="comment"> * Biases are 1D tensor with dimensions [OFM]. Data type supported: Same as @p input.</span></div><div class="line"><a name="l00301"></a><span class="lineno"> 301</span> <span class="comment"> * @param[in] output Output tensor info. Data type supported: Data type supported: QSYMM16</span></div><div class="line"><a name="l00302"></a><span class="lineno"> 302</span> <span class="comment"> * @param[in] min (Optional) Min value used to saturate down the output result before converting back to QSYMM16. Defaults to 0.</span></div><div class="line"><a name="l00303"></a><span class="lineno"> 303</span> <span class="comment"> * @param[in] max (Optional) Max value used to saturate up the output result before converting back to QSYMM16,</span></div><div class="line"><a name="l00304"></a><span class="lineno"> 304</span> <span class="comment"> * Along with @p min, this value can be used to implement "rectified linear unit" activation functions. Defaults to 0.</span></div><div class="line"><a name="l00305"></a><span class="lineno"> 305</span> <span class="comment"> *</span></div><div class="line"><a name="l00306"></a><span class="lineno"> 306</span> <span class="comment"> * @return a status</span></div><div class="line"><a name="l00307"></a><span class="lineno"> 307</span> <span class="comment"> */</span></div><div class="line"><a name="l00308"></a><span class="lineno"> 308</span>  <span class="keyword">static</span> <a class="code" href="classarm__compute_1_1_status.xhtml">Status</a> <a class="code" href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int16_scale_by_fixed_point.xhtml#aee63e7671cf04d15be2da1b83d90e61b">validate</a>(<span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a8fcf2ddd9a1d58b1b280f5c0aed71845">input</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *<a class="code" href="namespacearm__compute_1_1test_1_1validation.xhtml#a3a77be8aebd8e00522b32061d46ccdbd">bias</a>, <span class="keyword">const</span> <a class="code" href="classarm__compute_1_1_i_tensor_info.xhtml">ITensorInfo</a> *output, <span class="keywordtype">int</span> min = 0, <span class="keywordtype">int</span> max = 0);</div><div class="line"><a name="l00309"></a><span class="lineno"> 309</span> };</div><div class="line"><a name="l00310"></a><span class="lineno"> 310</span> } <span class="comment">// namespace arm_compute</span></div><div class="line"><a name="l00311"></a><span class="lineno"> 311</span> <span class="preprocessor">#endif </span><span class="comment">/*ARM_COMPUTE_CLGEMMLOWPOUTPUTSTAGE_H */</span><span class="preprocessor"></span></div><div class="ttc" id="src_2core_2_c_l_2cl__kernels_2_helpers_8h_xhtml_a009469e4d9b8fce3b6d5e97d2077827d"><div class="ttname"><a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h.xhtml#a009469e4d9b8fce3b6d5e97d2077827d">offset</a></div><div class="ttdeci">__global uchar * offset(const Image *img, int x, int y)</div><div class="ttdoc">Get the pointer position of a Image.</div><div class="ttdef"><b>Definition:</b> <a href="src_2core_2_c_l_2cl__kernels_2_helpers_8h_source.xhtml#l00510">helpers.h:510</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int8_scale_by_fixed_point_xhtml_a15da37f661fdcd81c1f25c5d6bdc6abd"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int8_scale_by_fixed_point.xhtml#a15da37f661fdcd81c1f25c5d6bdc6abd">arm_compute::CLGEMMLowpQuantizeDownInt32ToInt8ScaleByFixedPoint::configure</a></div><div class="ttdeci">void configure(const ICLTensor *input, const ICLTensor *bias, ICLTensor *output, int result_fixedpoint_multiplier, int result_shift, int result_offset_after_shift, int min=0, int max=0)</div><div class="ttdoc">Initialise the kernel's inputs, output.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8cpp_source.xhtml#l00063">CLGEMMLowpOutputStage.cpp:63</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int16_scale_by_fixed_point_xhtml_aee63e7671cf04d15be2da1b83d90e61b"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int16_scale_by_fixed_point.xhtml#aee63e7671cf04d15be2da1b83d90e61b">arm_compute::CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPoint::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *input, const ITensorInfo *bias, const ITensorInfo *output, int min=0, int max=0)</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMLowpQuantizeDownIn...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8cpp_source.xhtml#l00102">CLGEMMLowpOutputStage.cpp:102</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_fixed_point_xhtml_a15da37f661fdcd81c1f25c5d6bdc6abd"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_fixed_point.xhtml#a15da37f661fdcd81c1f25c5d6bdc6abd">arm_compute::CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint::configure</a></div><div class="ttdeci">void configure(const ICLTensor *input, const ICLTensor *bias, ICLTensor *output, int result_fixedpoint_multiplier, int result_shift, int result_offset_after_shift, int min=0, int max=0)</div><div class="ttdoc">Initialise the kernel's inputs, output.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8cpp_source.xhtml#l00048">CLGEMMLowpOutputStage.cpp:48</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_tensor_info_xhtml"><div class="ttname"><a href="classarm__compute_1_1_i_tensor_info.xhtml">arm_compute::ITensorInfo</a></div><div class="ttdoc">Store the tensor's metadata.</div><div class="ttdef"><b>Definition:</b> <a href="_i_tensor_info_8h_source.xhtml#l00040">ITensorInfo.h:40</a></div></div> |
| <div class="ttc" id="_i_c_l_simple_function_8h_xhtml"><div class="ttname"><a href="_i_c_l_simple_function_8h.xhtml">ICLSimpleFunction.h</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int16_scale_by_fixed_point_xhtml_a5c7abeeecc0f2d6c50cc1dafaadf7c19"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int16_scale_by_fixed_point.xhtml#a5c7abeeecc0f2d6c50cc1dafaadf7c19">arm_compute::CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPoint::configure</a></div><div class="ttdeci">void configure(const ICLTensor *input, const ICLTensor *bias, ICLTensor *output, int result_fixedpoint_multiplier, int result_shift, int min=0, int max=0)</div><div class="ttdoc">Initialise the kernel's inputs, output.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8cpp_source.xhtml#l00093">CLGEMMLowpOutputStage.cpp:93</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_status_xhtml"><div class="ttname"><a href="classarm__compute_1_1_status.xhtml">arm_compute::Status</a></div><div class="ttdoc">Status class.</div><div class="ttdef"><b>Definition:</b> <a href="_error_8h_source.xhtml#l00052">Error.h:52</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_float_xhtml_a49bf33836c645a5b5aa385411ab56964"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_float.xhtml#a49bf33836c645a5b5aa385411ab56964">arm_compute::CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloat::configure</a></div><div class="ttdeci">void configure(const ICLTensor *input, const ICLTensor *bias, ICLTensor *output, float multiplier, int offset, int min=0, int max=0)</div><div class="ttdoc">Initialise the kernel's inputs, output.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8cpp_source.xhtml#l00078">CLGEMMLowpOutputStage.cpp:78</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_xhtml_af127cda81674a7864000a3e70877c4c2"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale.xhtml#af127cda81674a7864000a3e70877c4c2">arm_compute::CLGEMMLowpQuantizeDownInt32ToUint8Scale::configure</a></div><div class="ttdeci">void configure(const ICLTensor *input, const ICLTensor *bias, ICLTensor *output, int result_offset, int result_mult_int, int result_shift, int min=0, int max=0)</div><div class="ttdoc">Initialise the kernel's inputs, output.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8cpp_source.xhtml#l00036">CLGEMMLowpOutputStage.cpp:36</a></div></div> |
| <div class="ttc" id="namespacearm__compute_xhtml"><div class="ttname"><a href="namespacearm__compute.xhtml">arm_compute</a></div><div class="ttdoc">Copyright (c) 2017-2020 ARM Limited.</div><div class="ttdef"><b>Definition:</b> <a href="00__introduction_8dox_source.xhtml#l00024">00_introduction.dox:24</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1test_1_1validation_xhtml_a8fcf2ddd9a1d58b1b280f5c0aed71845"><div class="ttname"><a href="namespacearm__compute_1_1test_1_1validation.xhtml#a8fcf2ddd9a1d58b1b280f5c0aed71845">arm_compute::test::validation::input</a></div><div class="ttdeci">auto input</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_2_l_s_t_m_layer_quantized_8cpp_source.xhtml#l00487">LSTMLayerQuantized.cpp:487</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_fixed_point_xhtml_aee63e7671cf04d15be2da1b83d90e61b"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_fixed_point.xhtml#aee63e7671cf04d15be2da1b83d90e61b">arm_compute::CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *input, const ITensorInfo *bias, const ITensorInfo *output, int min=0, int max=0)</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMLowpQuantizeDownIn...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8cpp_source.xhtml#l00057">CLGEMMLowpOutputStage.cpp:57</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_xhtml"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale.xhtml">arm_compute::CLGEMMLowpQuantizeDownInt32ToUint8Scale</a></div><div class="ttdoc">Basic function to execute CLGEMMLowpQuantizeDownInt32ToUint8Scale on OpenCL.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8h_source.xhtml#l00059">CLGEMMLowpOutputStage.h:59</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int8_scale_by_fixed_point_xhtml_aee63e7671cf04d15be2da1b83d90e61b"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int8_scale_by_fixed_point.xhtml#aee63e7671cf04d15be2da1b83d90e61b">arm_compute::CLGEMMLowpQuantizeDownInt32ToInt8ScaleByFixedPoint::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *input, const ITensorInfo *bias, const ITensorInfo *output, int min=0, int max=0)</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMLowpQuantizeDownIn...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8cpp_source.xhtml#l00072">CLGEMMLowpOutputStage.cpp:72</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_c_l_simple_function_xhtml"><div class="ttname"><a href="classarm__compute_1_1_i_c_l_simple_function.xhtml">arm_compute::ICLSimpleFunction</a></div><div class="ttdoc">Basic interface for functions which have a single OpenCL kernel.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_simple_function_8h_source.xhtml#l00039">ICLSimpleFunction.h:39</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int16_scale_by_fixed_point_xhtml"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int16_scale_by_fixed_point.xhtml">arm_compute::CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPoint</a></div><div class="ttdoc">Basic function to execute CLGEMMLowpQuantizeDownInt32ToInt16ScaleByFixedPoint on OpenCL.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8h_source.xhtml#l00280">CLGEMMLowpOutputStage.h:280</a></div></div> |
| <div class="ttc" id="namespacearm__compute_1_1test_1_1validation_xhtml_a3a77be8aebd8e00522b32061d46ccdbd"><div class="ttname"><a href="namespacearm__compute_1_1test_1_1validation.xhtml#a3a77be8aebd8e00522b32061d46ccdbd">arm_compute::test::validation::bias</a></div><div class="ttdeci">CLTensor bias</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_2_convolution_layer_8cpp_source.xhtml#l00189">ConvolutionLayer.cpp:189</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int8_scale_by_fixed_point_xhtml"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_int8_scale_by_fixed_point.xhtml">arm_compute::CLGEMMLowpQuantizeDownInt32ToInt8ScaleByFixedPoint</a></div><div class="ttdoc">Basic function to execute CLGEMMLowpQuantizeDownInt32ToInt8ScaleByFixedPoint on OpenCL.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8h_source.xhtml#l00180">CLGEMMLowpOutputStage.h:180</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_i_c_l_tensor_xhtml"><div class="ttname"><a href="classarm__compute_1_1_i_c_l_tensor.xhtml">arm_compute::ICLTensor</a></div><div class="ttdoc">Interface for OpenCL tensor.</div><div class="ttdef"><b>Definition:</b> <a href="_i_c_l_tensor_8h_source.xhtml#l00042">ICLTensor.h:42</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_float_xhtml"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_float.xhtml">arm_compute::CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloat</a></div><div class="ttdoc">Basic function to execute CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloat on OpenCL.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8h_source.xhtml#l00222">CLGEMMLowpOutputStage.h:222</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_float_xhtml_aee63e7671cf04d15be2da1b83d90e61b"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_float.xhtml#aee63e7671cf04d15be2da1b83d90e61b">arm_compute::CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFloat::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *input, const ITensorInfo *bias, const ITensorInfo *output, int min=0, int max=0)</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMLowpQuantizeDownIn...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8cpp_source.xhtml#l00087">CLGEMMLowpOutputStage.cpp:87</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_xhtml_aee63e7671cf04d15be2da1b83d90e61b"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale.xhtml#aee63e7671cf04d15be2da1b83d90e61b">arm_compute::CLGEMMLowpQuantizeDownInt32ToUint8Scale::validate</a></div><div class="ttdeci">static Status validate(const ITensorInfo *input, const ITensorInfo *bias, const ITensorInfo *output, int min=0, int max=0)</div><div class="ttdoc">Static function to check if given info will lead to a valid configuration of CLGEMMLowpQuantizeDownIn...</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8cpp_source.xhtml#l00043">CLGEMMLowpOutputStage.cpp:43</a></div></div> |
| <div class="ttc" id="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_fixed_point_xhtml"><div class="ttname"><a href="classarm__compute_1_1_c_l_g_e_m_m_lowp_quantize_down_int32_to_uint8_scale_by_fixed_point.xhtml">arm_compute::CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint</a></div><div class="ttdoc">Basic function to execute CLGEMMLowpQuantizeDownInt32ToUint8ScaleByFixedPoint on OpenCL.</div><div class="ttdef"><b>Definition:</b> <a href="_c_l_g_e_m_m_lowp_output_stage_8h_source.xhtml#l00119">CLGEMMLowpOutputStage.h:119</a></div></div> |
| </div><!-- fragment --></div><!-- contents --> |
| </div><!-- doc-content --> |
| <!-- start footer part --> |
| <div id="nav-path" class="navpath"><!-- id is needed for treeview function! --> |
| <ul> |
| <li class="navelem"><a class="el" href="dir_214608ef36d61442cb2b0c1c4e9a7def.xhtml">arm_compute</a></li><li class="navelem"><a class="el" href="dir_8bc26130589aa16388b5a02f17abf2c2.xhtml">runtime</a></li><li class="navelem"><a class="el" href="dir_b986aa98f9b6490bb4fc7918500346fb.xhtml">CL</a></li><li class="navelem"><a class="el" href="dir_251c63bc3f70b60acf43edd6d870a3ef.xhtml">functions</a></li><li class="navelem"><a class="el" href="_c_l_g_e_m_m_lowp_output_stage_8h.xhtml">CLGEMMLowpOutputStage.h</a></li> |
| <li class="footer">Generated on Thu Mar 5 2020 16:06:56 for Compute Library by |
| <a href="http://www.doxygen.org/index.html"> |
| <img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.8.15 </li> |
| </ul> |
| </div> |
| </body> |
| </html> |
