diff options
| author | Nikhil Raj <nikhil.raj@arm.com> | 2021-11-17 13:16:45 +0000 |
|---|---|---|
| committer | Nikhil Raj <nikhil.raj@arm.com> | 2021-11-17 13:16:45 +0000 |
| commit | 9aed8fb43441228343b925b42464a55042c47ca0 (patch) | |
| tree | 4c34534eea1c8e82655ac1f60e3633b9618cc40d /21.11/_delegate_quick_start_guide_8md_source.xhtml | |
| parent | f86be93b7492b381370cae7bf71eca8572a0cbae (diff) | |
| download | armnn-9aed8fb43441228343b925b42464a55042c47ca0.tar.gz | |
IVGCVSW-6040 Update 21.11 Doxygen Documents
Signed-off-by: Nikhil Raj <nikhil.raj@arm.com>
Change-Id: Ia36ec98c4bebc27a69103911ea3409cd7db587a5
Diffstat (limited to '21.11/_delegate_quick_start_guide_8md_source.xhtml')
| -rw-r--r-- | 21.11/_delegate_quick_start_guide_8md_source.xhtml | 113 |
1 files changed, 113 insertions, 0 deletions
diff --git a/21.11/_delegate_quick_start_guide_8md_source.xhtml b/21.11/_delegate_quick_start_guide_8md_source.xhtml new file mode 100644 index 0000000000..d86d33335b --- /dev/null +++ b/21.11/_delegate_quick_start_guide_8md_source.xhtml @@ -0,0 +1,113 @@ +<!-- Copyright (c) 2020 ARM Limited. --> +<!-- --> +<!-- SPDX-License-Identifier: MIT --> +<!-- --> +<!-- HTML header for doxygen 1.8.13--> +<!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 1.8.13"/> +<meta name="robots" content="NOINDEX, NOFOLLOW" /> +<meta name="viewport" content="width=device-width, initial-scale=1"/> +<title>ArmNN: delegate/DelegateQuickStartGuide.md 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"> + $(document).ready(initResizable); +</script> +<link href="search/search.css" rel="stylesheet" type="text/css"/> +<script type="text/javascript" src="search/searchdata.js"></script> +<script type="text/javascript" src="search/search.js"></script> +<script type="text/x-mathjax-config"> + MathJax.Hub.Config({ + extensions: ["tex2jax.js"], + jax: ["input/TeX","output/HTML-CSS"], +}); +</script><script type="text/javascript" src="http://cdn.mathjax.org/mathjax/latest/MathJax.js"></script> +<link href="doxygen.css" rel="stylesheet" type="text/css" /> +<link href="stylesheet.css" rel="stylesheet" type="text/css"/> +</head> +<body> +<div id="top"><!-- do not remove this div, it is closed by doxygen! --> +<div id="titlearea"> +<table cellspacing="0" cellpadding="0"> + <tbody> + <tr style="height: 56px;"> + <img alt="ArmNN" src="Arm_NN_horizontal_blue.png" style="max-width: 10rem; margin-top: .5rem; margin-left 10px"/> + <td style="padding-left: 0.5em;"> + <div id="projectname"> +  <span id="projectnumber">21.11</span> + </div> + </td> + </tr> + </tbody> +</table> +</div> +<!-- end header part --> +<!-- Generated by Doxygen 1.8.13 --> +<script type="text/javascript"> +var searchBox = new SearchBox("searchBox", "search",false,'Search'); +</script> +<script type="text/javascript" src="menudata.js"></script> +<script type="text/javascript" src="menu.js"></script> +<script type="text/javascript"> +$(function() { + initMenu('',true,false,'search.php','Search'); + $(document).ready(function() { init_search(); }); +}); +</script> +<div id="main-nav"></div> +</div><!-- top --> +<div id="side-nav" class="ui-resizable side-nav-resizable"> + <div id="nav-tree"> + <div id="nav-tree-contents"> + <div id="nav-sync" class="sync"></div> + </div> + </div> + <div id="splitbar" style="-moz-user-select:none;" + class="ui-resizable-handle"> + </div> +</div> +<script type="text/javascript"> +$(document).ready(function(){initNavTree('_delegate_quick_start_guide_8md.xhtml','');}); +</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">delegate/DelegateQuickStartGuide.md</div> </div> +</div><!--header--> +<div class="contents"> +<a href="_delegate_quick_start_guide_8md.xhtml">Go to the documentation of this file.</a><div class="fragment"><div class="line"><a name="l00001"></a><span class="lineno"> 1</span> # TfLite Delegate Quick Start Guide</div><div class="line"><a name="l00002"></a><span class="lineno"> 2</span> If you have downloaded the ArmNN Github binaries or built the TfLite delegate yourself, then this tutorial will show you how you can</div><div class="line"><a name="l00003"></a><span class="lineno"> 3</span> integrate it into TfLite to run models using python.</div><div class="line"><a name="l00004"></a><span class="lineno"> 4</span> </div><div class="line"><a name="l00005"></a><span class="lineno"> 5</span> Here is an example python script showing how to do this. In this script we are making use of the </div><div class="line"><a name="l00006"></a><span class="lineno"> 6</span> [external adaptor](https://www.tensorflow.org/lite/performance/implementing_delegate#option_2_leverage_external_delegate) </div><div class="line"><a name="l00007"></a><span class="lineno"> 7</span> tool of TfLite that allows you to load delegates at runtime.</div><div class="line"><a name="l00008"></a><span class="lineno"> 8</span> ```python</div><div class="line"><a name="l00009"></a><span class="lineno"> 9</span> import numpy as np</div><div class="line"><a name="l00010"></a><span class="lineno"> 10</span> import tflite_runtime.interpreter as tflite</div><div class="line"><a name="l00011"></a><span class="lineno"> 11</span> </div><div class="line"><a name="l00012"></a><span class="lineno"> 12</span> # Load TFLite model and allocate tensors.</div><div class="line"><a name="l00013"></a><span class="lineno"> 13</span> # (if you are using the complete tensorflow package you can find load_delegate in tf.experimental.load_delegate)</div><div class="line"><a name="l00014"></a><span class="lineno"> 14</span> armnn_delegate = tflite.load_delegate( library="<path-to-armnn-binaries>/libarmnnDelegate.so",</div><div class="line"><a name="l00015"></a><span class="lineno"> 15</span>  options={"backends": "CpuAcc,GpuAcc,CpuRef", "logging-severity":"info"})</div><div class="line"><a name="l00016"></a><span class="lineno"> 16</span> # Delegates/Executes all operations supported by ArmNN to/with ArmNN</div><div class="line"><a name="l00017"></a><span class="lineno"> 17</span> interpreter = tflite.Interpreter(model_path="<your-armnn-repo-dir>/delegate/python/test/test_data/mock_model.tflite", </div><div class="line"><a name="l00018"></a><span class="lineno"> 18</span>  experimental_delegates=[armnn_delegate])</div><div class="line"><a name="l00019"></a><span class="lineno"> 19</span> interpreter.allocate_tensors()</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> # Get input and output tensors.</div><div class="line"><a name="l00022"></a><span class="lineno"> 22</span> input_details = interpreter.get_input_details()</div><div class="line"><a name="l00023"></a><span class="lineno"> 23</span> output_details = interpreter.get_output_details()</div><div class="line"><a name="l00024"></a><span class="lineno"> 24</span> </div><div class="line"><a name="l00025"></a><span class="lineno"> 25</span> # Test model on random input data.</div><div class="line"><a name="l00026"></a><span class="lineno"> 26</span> input_shape = input_details[0]['shape']</div><div class="line"><a name="l00027"></a><span class="lineno"> 27</span> input_data = np.array(np.random.random_sample(input_shape), dtype=np.uint8)</div><div class="line"><a name="l00028"></a><span class="lineno"> 28</span> interpreter.set_tensor(input_details[0]['index'], input_data)</div><div class="line"><a name="l00029"></a><span class="lineno"> 29</span> </div><div class="line"><a name="l00030"></a><span class="lineno"> 30</span> interpreter.invoke()</div><div class="line"><a name="l00031"></a><span class="lineno"> 31</span> </div><div class="line"><a name="l00032"></a><span class="lineno"> 32</span> # Print out result</div><div class="line"><a name="l00033"></a><span class="lineno"> 33</span> output_data = interpreter.get_tensor(output_details[0]['index'])</div><div class="line"><a name="l00034"></a><span class="lineno"> 34</span> print(output_data)</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> </div><div class="line"><a name="l00037"></a><span class="lineno"> 37</span> # Prepare the environment</div><div class="line"><a name="l00038"></a><span class="lineno"> 38</span> Pre-requisites:</div><div class="line"><a name="l00039"></a><span class="lineno"> 39</span>  * Dynamically build Arm NN Delegate library or download the ArmNN binaries</div><div class="line"><a name="l00040"></a><span class="lineno"> 40</span>  * python3 (Depends on TfLite version)</div><div class="line"><a name="l00041"></a><span class="lineno"> 41</span>  * virtualenv</div><div class="line"><a name="l00042"></a><span class="lineno"> 42</span>  * numpy (Depends on TfLite version)</div><div class="line"><a name="l00043"></a><span class="lineno"> 43</span>  * tflite_runtime (>=2.5, depends on Arm NN Delegate)</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> If you haven't built the delegate yet then take a look at the [build guide](./BuildGuideNative.md). Otherwise, </div><div class="line"><a name="l00046"></a><span class="lineno"> 46</span> you can download the binaries [here](https://github.com/ARM-software/armnn/releases/tag/v21.11)</div><div class="line"><a name="l00047"></a><span class="lineno"> 47</span> </div><div class="line"><a name="l00048"></a><span class="lineno"> 48</span> We recommend creating a virtual environment for this tutorial. For the following code to work python3 is needed. Please</div><div class="line"><a name="l00049"></a><span class="lineno"> 49</span> also check the documentation of the TfLite version you want to use. There might be additional prerequisites for the python</div><div class="line"><a name="l00050"></a><span class="lineno"> 50</span> version. We will use Tensorflow Lite 2.5.1 for this guide.</div><div class="line"><a name="l00051"></a><span class="lineno"> 51</span> ```bash</div><div class="line"><a name="l00052"></a><span class="lineno"> 52</span> # Install python3 (We ended up with python3.5.3) and virtualenv</div><div class="line"><a name="l00053"></a><span class="lineno"> 53</span> sudo apt-get install python3-pip</div><div class="line"><a name="l00054"></a><span class="lineno"> 54</span> sudo pip3 install virtualenv</div><div class="line"><a name="l00055"></a><span class="lineno"> 55</span> </div><div class="line"><a name="l00056"></a><span class="lineno"> 56</span> # create a virtual environment</div><div class="line"><a name="l00057"></a><span class="lineno"> 57</span> cd your/tutorial/dir</div><div class="line"><a name="l00058"></a><span class="lineno"> 58</span> # creates a directory myenv at the current location</div><div class="line"><a name="l00059"></a><span class="lineno"> 59</span> virtualenv -p python3 myenv </div><div class="line"><a name="l00060"></a><span class="lineno"> 60</span> # activate the environment</div><div class="line"><a name="l00061"></a><span class="lineno"> 61</span> source myenv/bin/activate</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> Now that the environment is active we can install additional packages we need for our example script. As you can see </div><div class="line"><a name="l00065"></a><span class="lineno"> 65</span> in the python script at the start of this page, this tutorial uses the `tflite_runtime` rather than the whole tensorflow </div><div class="line"><a name="l00066"></a><span class="lineno"> 66</span> package. The `tflite_runtime` is a package that wraps the TfLite Interpreter. Therefore it can only be used to run inferences of </div><div class="line"><a name="l00067"></a><span class="lineno"> 67</span> TfLite models. But since Arm NN is only an inference engine itself this is a perfect match. The </div><div class="line"><a name="l00068"></a><span class="lineno"> 68</span> `tflite_runtime` is also much smaller than the whole tensorflow package and better suited to run models on </div><div class="line"><a name="l00069"></a><span class="lineno"> 69</span> mobile and embedded devices.</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> The TfLite [website](https://www.tensorflow.org/lite/guide/python) shows you two methods to download the `tflite_runtime` package. </div><div class="line"><a name="l00072"></a><span class="lineno"> 72</span> In our experience, the use of the pip command works for most systems including debian. However, if you're using an older version of Tensorflow, </div><div class="line"><a name="l00073"></a><span class="lineno"> 73</span> you may need to build the pip package from source. You can find more information [here](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/tools/pip_package/README.md).</div><div class="line"><a name="l00074"></a><span class="lineno"> 74</span> But in our case, with Tensorflow Lite 2.5.1, we can install through:</div><div class="line"><a name="l00075"></a><span class="lineno"> 75</span> </div><div class="line"><a name="l00076"></a><span class="lineno"> 76</span> ```</div><div class="line"><a name="l00077"></a><span class="lineno"> 77</span> pip3 install --extra-index-url https://google-coral.github.io/py-repo/ tflite_runtime</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> </div><div class="line"><a name="l00080"></a><span class="lineno"> 80</span> Your virtual environment is now all setup. Copy the final python script into a python file e.g. </div><div class="line"><a name="l00081"></a><span class="lineno"> 81</span> `ExternalDelegatePythonTutorial.py`. Modify the python script above and replace `<path-to-armnn-binaries>` and </div><div class="line"><a name="l00082"></a><span class="lineno"> 82</span> `<your-armnn-repo-dir>` with the directories you have set up. If you've been using the [native build guide](./BuildGuideNative.md) </div><div class="line"><a name="l00083"></a><span class="lineno"> 83</span> this will be `$BASEDIR/armnn/build` and `$BASEDIR/armnn`.</div><div class="line"><a name="l00084"></a><span class="lineno"> 84</span> </div><div class="line"><a name="l00085"></a><span class="lineno"> 85</span> Finally, execute the script:</div><div class="line"><a name="l00086"></a><span class="lineno"> 86</span> ```bash</div><div class="line"><a name="l00087"></a><span class="lineno"> 87</span> python ExternalDelegatePythonTutorial.py</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> The output should look similar to this:</div><div class="line"><a name="l00090"></a><span class="lineno"> 90</span> ```bash</div><div class="line"><a name="l00091"></a><span class="lineno"> 91</span> Info: ArmNN v27.0.0</div><div class="line"><a name="l00092"></a><span class="lineno"> 92</span> </div><div class="line"><a name="l00093"></a><span class="lineno"> 93</span> Info: Initialization time: 0.56 ms</div><div class="line"><a name="l00094"></a><span class="lineno"> 94</span> </div><div class="line"><a name="l00095"></a><span class="lineno"> 95</span> INFO: TfLiteArmnnDelegate: Created TfLite ArmNN delegate.</div><div class="line"><a name="l00096"></a><span class="lineno"> 96</span> [[ 12 123 16 12 11 14 20 16 20 12]]</div><div class="line"><a name="l00097"></a><span class="lineno"> 97</span> Info: Shutdown time: 0.28 ms</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> </div><div class="line"><a name="l00100"></a><span class="lineno"> 100</span> For more details of the kind of options you can pass to the Arm NN delegate please check the parameters of function [tflite_plugin_create_delegate](namespacetflite.xhtml).</div><div class="line"><a name="l00101"></a><span class="lineno"> 101</span> </div><div class="line"><a name="l00102"></a><span class="lineno"> 102</span> You can also test the functionality of the external delegate adaptor by running some unit tests:</div><div class="line"><a name="l00103"></a><span class="lineno"> 103</span> ```bash</div><div class="line"><a name="l00104"></a><span class="lineno"> 104</span> pip install pytest</div><div class="line"><a name="l00105"></a><span class="lineno"> 105</span> cd armnn/delegate/python/test</div><div class="line"><a name="l00106"></a><span class="lineno"> 106</span> # You can deselect tests that require backends that your hardware doesn't support using markers e.g. -m "not GpuAccTest"</div><div class="line"><a name="l00107"></a><span class="lineno"> 107</span> pytest --delegate-dir="<path-to-armnn-binaries>/libarmnnDelegate.so" -m "not GpuAccTest"</div><div class="line"><a name="l00108"></a><span class="lineno"> 108</span> ```</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="_delegate_quick_start_guide_8md.xhtml">DelegateQuickStartGuide.md</a></li> + <li class="footer">Generated on Wed Nov 17 2021 12:58:59 for ArmNN by + <a href="http://www.doxygen.org/index.html"> + <img class="footer" src="doxygen.png" alt="doxygen"/></a> 1.8.13 </li> + </ul> +</div> +</body> +</html> |