diff options
Diffstat (limited to '21.02/_r_e_a_d_m_e_8md_source.xhtml')
-rw-r--r-- | 21.02/_r_e_a_d_m_e_8md_source.xhtml | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/21.02/_r_e_a_d_m_e_8md_source.xhtml b/21.02/_r_e_a_d_m_e_8md_source.xhtml index 009ba42910..0f1377c19b 100644 --- a/21.02/_r_e_a_d_m_e_8md_source.xhtml +++ b/21.02/_r_e_a_d_m_e_8md_source.xhtml @@ -98,13 +98,13 @@ $(document).ready(function(){initNavTree('_r_e_a_d_m_e_8md.xhtml','');}); <div class="title">README.md</div> </div> </div><!--header--> <div class="contents"> -<a href="_r_e_a_d_m_e_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> # Introduction</div><div class="line"><a name="l00002"></a><span class="lineno"> 2</span> </div><div class="line"><a name="l00003"></a><span class="lineno"> 3</span> * [Software tools overview](#software-tools-overview)</div><div class="line"><a name="l00004"></a><span class="lineno"> 4</span> * [Where to find more information](#where-to-find-more-information)</div><div class="line"><a name="l00005"></a><span class="lineno"> 5</span> * [Contributions](#contributions)</div><div class="line"><a name="l00006"></a><span class="lineno"> 6</span> * [Disclaimer](#disclaimer)</div><div class="line"><a name="l00007"></a><span class="lineno"> 7</span> * [License](#license)</div><div class="line"><a name="l00008"></a><span class="lineno"> 8</span> * [Third-Party](#third-party)</div><div class="line"><a name="l00009"></a><span class="lineno"> 9</span> </div><div class="line"><a name="l00010"></a><span class="lineno"> 10</span> Arm NN is a key component of the [machine learning platform](https://mlplatform.org/), which is part of the </div><div class="line"><a name="l00011"></a><span class="lineno"> 11</span> [Linaro Machine Intelligence Initiative](https://www.linaro.org/news/linaro-announces-launch-of-machine-intelligence-initiative/). </div><div class="line"><a name="l00012"></a><span class="lineno"> 12</span> </div><div class="line"><a name="l00013"></a><span class="lineno"> 13</span> The Arm NN SDK is a set of open-source software and tools that enables machine learning workloads on power-efficient </div><div class="line"><a name="l00014"></a><span class="lineno"> 14</span> devices. It provides a bridge between existing neural network frameworks and power-efficient Cortex-A CPUs, </div><div class="line"><a name="l00015"></a><span class="lineno"> 15</span> Arm Mali GPUs and Arm Ethos NPUs.</div><div class="line"><a name="l00016"></a><span class="lineno"> 16</span> </div><div class="line"><a name="l00017"></a><span class="lineno"> 17</span> <img align="center" width="400" src="https://developer.arm.com/-/media/Arm Developer Community/Images/Block Diagrams/Arm-NN/Arm-NN-Frameworks-Diagram.png"/></div><div class="line"><a name="l00018"></a><span class="lineno"> 18</span> </div><div class="line"><a name="l00019"></a><span class="lineno"> 19</span> Arm NN SDK utilizes the Compute Library to target programmable cores, such as Cortex-A CPUs and Mali GPUs, </div><div class="line"><a name="l00020"></a><span class="lineno"> 20</span> as efficiently as possible. To target Ethos NPUs the NPU-Driver is utilized. We also welcome new contributors to provide </div><div class="line"><a name="l00021"></a><span class="lineno"> 21</span> their [own driver and backend](src/backends/README.md). Note, Arm NN does not provide support for Cortex-M CPUs.</div><div class="line"><a name="l00022"></a><span class="lineno"> 22</span> </div><div class="line"><a name="l00023"></a><span class="lineno"> 23</span> The latest release supports models created with **Caffe**, **TensorFlow**, **TensorFlow Lite** (TfLite) and **ONNX**. </div><div class="line"><a name="l00024"></a><span class="lineno"> 24</span> Arm NN analysis a given model and replaces the operations within it with implementations particularly designed for the </div><div class="line"><a name="l00025"></a><span class="lineno"> 25</span> hardware you want to execute it on. This results in a great boost of execution speed. How much faster your neural </div><div class="line"><a name="l00026"></a><span class="lineno"> 26</span> network can be executed depends on the operations it contains and the available hardware. Below you can see the speedup </div><div class="line"><a name="l00027"></a><span class="lineno"> 27</span> we've been experiencing in our experiments with a few common networks.</div><div class="line"><a name="l00028"></a><span class="lineno"> 28</span> </div><div class="line"><a name="l00029"></a><span class="lineno"> 29</span> <img align="center" width="700" src="https://developer.arm.com/-/media/developer/Other Images/Arm_NN_performance_relative_to_other_NN_frameworks_diagram.png"/></div><div class="line"><a name="l00030"></a><span class="lineno"> 30</span> </div><div class="line"><a name="l00031"></a><span class="lineno"> 31</span> Arm NN is written using portable C++14 and the build system uses [CMake](https://cmake.org/), therefore it is possible </div><div class="line"><a name="l00032"></a><span class="lineno"> 32</span> to build for a wide variety of target platforms, from a wide variety of host environments.</div><div class="line"><a name="l00033"></a><span class="lineno"> 33</span> </div><div class="line"><a name="l00034"></a><span class="lineno"> 34</span> </div><div class="line"><a name="l00035"></a><span class="lineno"> 35</span> ## Getting started: Software tools overview</div><div class="line"><a name="l00036"></a><span class="lineno"> 36</span> Depending on what kind of framework (Tensorflow, Caffe, ONNX) you've been using to create your model there are multiple </div><div class="line"><a name="l00037"></a><span class="lineno"> 37</span> software tools available within Arm NN that can serve your needs.</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> Generally, there is a **parser** available **for each supported framework**. Each parser allows you to run models from </div><div class="line"><a name="l00040"></a><span class="lineno"> 40</span> one framework e.g. the TfLite-Parser lets you run TfLite models. You can integrate these parsers into your own </div><div class="line"><a name="l00041"></a><span class="lineno"> 41</span> application to load, optimize and execute your model. We also provide **python bindings** for our parsers and the Arm NN core.</div><div class="line"><a name="l00042"></a><span class="lineno"> 42</span> We call the result **PyArmNN**. Therefore your application can be conveniently written in either C++ using the "original"</div><div class="line"><a name="l00043"></a><span class="lineno"> 43</span> Arm NN library or in Python using PyArmNN. You can find tutorials on how to setup and use our parsers in our doxygen</div><div class="line"><a name="l00044"></a><span class="lineno"> 44</span> documentation. The latest version can be found in the [wiki section](https://github.com/ARM-software/armnn/wiki/Documentation) </div><div class="line"><a name="l00045"></a><span class="lineno"> 45</span> of this repository.</div><div class="line"><a name="l00046"></a><span class="lineno"> 46</span> </div><div class="line"><a name="l00047"></a><span class="lineno"> 47</span> Admittedly, building Arm NN and its parsers from source is not always easy to accomplish. We are trying to increase our</div><div class="line"><a name="l00048"></a><span class="lineno"> 48</span> usability by providing Arm NN as a **Debian package**. Our debian package is the most easy way to install the Arm NN Core,</div><div class="line"><a name="l00049"></a><span class="lineno"> 49</span> the TfLite Parser and PyArmNN (More support is about to come): [Installation via Apt Repository](InstallationViaAptRepository.md)</div><div class="line"><a name="l00050"></a><span class="lineno"> 50</span> </div><div class="line"><a name="l00051"></a><span class="lineno"> 51</span> The newest member in Arm NNs software toolkit is the **TfLite Delegate**. The delegate can be integrated in TfLite.</div><div class="line"><a name="l00052"></a><span class="lineno"> 52</span> TfLite will then delegate operations, that can be accelerated with Arm NN, to Arm NN. Every other operation will still be</div><div class="line"><a name="l00053"></a><span class="lineno"> 53</span> executed with the usual TfLite runtime. This is our **recommended way to accelerate TfLite models**. As with our parsers</div><div class="line"><a name="l00054"></a><span class="lineno"> 54</span> there are tutorials in our doxygen documentation that can be found in the [wiki section](https://github.com/ARM-software/armnn/wiki/Documentation).</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> If you would like to use **Arm NN on Android** you can follow this guide which explains [how to build Arm NN using the AndroidNDK]().</div><div class="line"><a name="l00057"></a><span class="lineno"> 57</span> But you might also want to take a look at another repository which implements a hardware abstraction layer (HAL) for </div><div class="line"><a name="l00058"></a><span class="lineno"> 58</span> Android. The repository is called [Android-NN-Driver](https://github.com/ARM-software/android-nn-driver) and when </div><div class="line"><a name="l00059"></a><span class="lineno"> 59</span> integrated into Android it will automatically run neural networks with Arm NN.</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> </div><div class="line"><a name="l00062"></a><span class="lineno"> 62</span> ## Where to find more information</div><div class="line"><a name="l00063"></a><span class="lineno"> 63</span> The section above introduces the most important tools that Arm NN provides.</div><div class="line"><a name="l00064"></a><span class="lineno"> 64</span> You can find a complete list in our **doxygen documentation**. The </div><div class="line"><a name="l00065"></a><span class="lineno"> 65</span> latest version can be found in the [wiki section](https://github.com/ARM-software/armnn/wiki/Documentation) of our github </div><div class="line"><a name="l00066"></a><span class="lineno"> 66</span> repository.</div><div class="line"><a name="l00067"></a><span class="lineno"> 67</span> </div><div class="line"><a name="l00068"></a><span class="lineno"> 68</span> For FAQs and troubleshooting advice, see [FAQ.md](docs/FAQ.md) </div><div class="line"><a name="l00069"></a><span class="lineno"> 69</span> or take a look at previous [github issues](https://github.com/ARM-software/armnn/issues).</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> </div><div class="line"><a name="l00072"></a><span class="lineno"> 72</span> ## Note</div><div class="line"><a name="l00073"></a><span class="lineno"> 73</span> 1. The following tools are **deprecated** in Arm NN 21.02 and will be removed in 21.05:</div><div class="line"><a name="l00074"></a><span class="lineno"> 74</span>  * TensorflowParser</div><div class="line"><a name="l00075"></a><span class="lineno"> 75</span>  * CaffeParser</div><div class="line"><a name="l00076"></a><span class="lineno"> 76</span>  * Quantizer</div><div class="line"><a name="l00077"></a><span class="lineno"> 77</span> </div><div class="line"><a name="l00078"></a><span class="lineno"> 78</span> 2. We are currently in the process of removing [boost](https://www.boost.org/) as a dependency to Arm NN. This process </div><div class="line"><a name="l00079"></a><span class="lineno"> 79</span> is finished for everything apart from our unit tests. This means you don't need boost to build and use Arm NN but </div><div class="line"><a name="l00080"></a><span class="lineno"> 80</span> you need it to execute our unit tests. Boost will soon be removed from Arm NN entirely.</div><div class="line"><a name="l00081"></a><span class="lineno"> 81</span> </div><div class="line"><a name="l00082"></a><span class="lineno"> 82</span> </div><div class="line"><a name="l00083"></a><span class="lineno"> 83</span> ## Contributions</div><div class="line"><a name="l00084"></a><span class="lineno"> 84</span> The Arm NN project welcomes contributions. For more details on contributing to Arm NN see the [Contributing page](https://mlplatform.org/contributing/) </div><div class="line"><a name="l00085"></a><span class="lineno"> 85</span> on the [MLPlatform.org](https://mlplatform.org/) website, or see the [Contributor Guide](ContributorGuide.md).</div><div class="line"><a name="l00086"></a><span class="lineno"> 86</span> </div><div class="line"><a name="l00087"></a><span class="lineno"> 87</span> Particularly if you'd like to implement your own backend next to our CPU, GPU and NPU backends there are guides for </div><div class="line"><a name="l00088"></a><span class="lineno"> 88</span> backend development: </div><div class="line"><a name="l00089"></a><span class="lineno"> 89</span> [Backend development guide](src/backends/README.md), [Dynamic backend development guide](src/dynamic/README.md)</div><div class="line"><a name="l00090"></a><span class="lineno"> 90</span> </div><div class="line"><a name="l00091"></a><span class="lineno"> 91</span> </div><div class="line"><a name="l00092"></a><span class="lineno"> 92</span> ## Disclaimer</div><div class="line"><a name="l00093"></a><span class="lineno"> 93</span> The armnn/tests directory contains tests used during Arm NN development. Many of them depend on third-party IP, model </div><div class="line"><a name="l00094"></a><span class="lineno"> 94</span> protobufs and image files not distributed with Arm NN. The dependencies of some of the tests are available freely on </div><div class="line"><a name="l00095"></a><span class="lineno"> 95</span> the Internet, for those who wish to experiment, but they won't run out of the box.</div><div class="line"><a name="l00096"></a><span class="lineno"> 96</span> </div><div class="line"><a name="l00097"></a><span class="lineno"> 97</span> </div><div class="line"><a name="l00098"></a><span class="lineno"> 98</span> ## License</div><div class="line"><a name="l00099"></a><span class="lineno"> 99</span> Arm NN is provided under the [MIT](https://spdx.org/licenses/MIT.html) license.</div><div class="line"><a name="l00100"></a><span class="lineno"> 100</span> See [LICENSE](LICENSE) for more information. Contributions to this project are accepted under the same license.</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> Individual files contain the following tag instead of the full license text.</div><div class="line"><a name="l00103"></a><span class="lineno"> 103</span> </div><div class="line"><a name="l00104"></a><span class="lineno"> 104</span>  SPDX-License-Identifier: MIT</div><div class="line"><a name="l00105"></a><span class="lineno"> 105</span> </div><div class="line"><a name="l00106"></a><span class="lineno"> 106</span> This enables machine processing of license information based on the SPDX License Identifiers that are available here: http://spdx.org/licenses/</div><div class="line"><a name="l00107"></a><span class="lineno"> 107</span> </div><div class="line"><a name="l00108"></a><span class="lineno"> 108</span> </div><div class="line"><a name="l00109"></a><span class="lineno"> 109</span> ## Third-party</div><div class="line"><a name="l00110"></a><span class="lineno"> 110</span> Third party tools used by Arm NN:</div><div class="line"><a name="l00111"></a><span class="lineno"> 111</span> </div><div class="line"><a name="l00112"></a><span class="lineno"> 112</span> | Tool | License (SPDX ID) | Description | Version | Provenience</div><div class="line"><a name="l00113"></a><span class="lineno"> 113</span> |----------------|-------------------|------------------------------------------------------------------|-------------|-------------------</div><div class="line"><a name="l00114"></a><span class="lineno"> 114</span> | cxxopts | MIT | A lightweight C++ option parser library | SHA 12e496da3d486b87fa9df43edea65232ed852510 | https://github.com/jarro2783/cxxopts</div><div class="line"><a name="l00115"></a><span class="lineno"> 115</span> | fmt | MIT | {fmt} is an open-source formatting library providing a fast and safe alternative to C stdio and C++ iostreams. | 7.0.1 | https://github.com/fmtlib/fmt</div><div class="line"><a name="l00116"></a><span class="lineno"> 116</span> | ghc | MIT | A header-only single-file std::filesystem compatible helper library | 1.3.2 | https://github.com/gulrak/filesystem</div><div class="line"><a name="l00117"></a><span class="lineno"> 117</span> | half | MIT | IEEE 754 conformant 16-bit half-precision floating point library | 1.12.0 | http://half.sourceforge.net</div><div class="line"><a name="l00118"></a><span class="lineno"> 118</span> | mapbox/variant | BSD | A header-only alternative to 'boost::variant' | 1.1.3 | https://github.com/mapbox/variant</div><div class="line"><a name="l00119"></a><span class="lineno"> 119</span> | stb | MIT | Image loader, resize and writer | 2.16 | https://github.com/nothings/stb</div></div><!-- fragment --></div><!-- contents --> +<a href="_r_e_a_d_m_e_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> # Introduction</div><div class="line"><a name="l00002"></a><span class="lineno"> 2</span> </div><div class="line"><a name="l00003"></a><span class="lineno"> 3</span> * [Software tools overview](#software-tools-overview)</div><div class="line"><a name="l00004"></a><span class="lineno"> 4</span> * [Where to find more information](#where-to-find-more-information)</div><div class="line"><a name="l00005"></a><span class="lineno"> 5</span> * [Contributions](#contributions)</div><div class="line"><a name="l00006"></a><span class="lineno"> 6</span> * [Disclaimer](#disclaimer)</div><div class="line"><a name="l00007"></a><span class="lineno"> 7</span> * [License](#license)</div><div class="line"><a name="l00008"></a><span class="lineno"> 8</span> * [Third-Party](#third-party)</div><div class="line"><a name="l00009"></a><span class="lineno"> 9</span> </div><div class="line"><a name="l00010"></a><span class="lineno"> 10</span> Arm NN is a key component of the [machine learning platform](https://mlplatform.org/), which is part of the</div><div class="line"><a name="l00011"></a><span class="lineno"> 11</span> [Linaro Machine Intelligence Initiative](https://www.linaro.org/news/linaro-announces-launch-of-machine-intelligence-initiative/).</div><div class="line"><a name="l00012"></a><span class="lineno"> 12</span> </div><div class="line"><a name="l00013"></a><span class="lineno"> 13</span> The Arm NN SDK is a set of open-source software and tools that enables machine learning workloads on power-efficient</div><div class="line"><a name="l00014"></a><span class="lineno"> 14</span> devices. It provides a bridge between existing neural network frameworks and power-efficient Cortex-A CPUs,</div><div class="line"><a name="l00015"></a><span class="lineno"> 15</span> Arm Mali GPUs and Arm Ethos NPUs.</div><div class="line"><a name="l00016"></a><span class="lineno"> 16</span> </div><div class="line"><a name="l00017"></a><span class="lineno"> 17</span> <img align="center" width="400" src="https://developer.arm.com/-/media/Arm Developer Community/Images/Block Diagrams/Arm-NN/Arm-NN-Frameworks-Diagram.png"/></div><div class="line"><a name="l00018"></a><span class="lineno"> 18</span> </div><div class="line"><a name="l00019"></a><span class="lineno"> 19</span> Arm NN SDK utilizes the Compute Library to target programmable cores, such as Cortex-A CPUs and Mali GPUs,</div><div class="line"><a name="l00020"></a><span class="lineno"> 20</span> as efficiently as possible. To target Ethos NPUs the NPU-Driver is utilized. We also welcome new contributors to provide</div><div class="line"><a name="l00021"></a><span class="lineno"> 21</span> their [own driver and backend](src/backends/README.md). Note, Arm NN does not provide support for Cortex-M CPUs.</div><div class="line"><a name="l00022"></a><span class="lineno"> 22</span> </div><div class="line"><a name="l00023"></a><span class="lineno"> 23</span> The latest release supports models created with **Caffe**, **TensorFlow**, **TensorFlow Lite** (TfLite) and **ONNX**.</div><div class="line"><a name="l00024"></a><span class="lineno"> 24</span> Arm NN analysis a given model and replaces the operations within it with implementations particularly designed for the</div><div class="line"><a name="l00025"></a><span class="lineno"> 25</span> hardware you want to execute it on. This results in a great boost of execution speed. How much faster your neural</div><div class="line"><a name="l00026"></a><span class="lineno"> 26</span> network can be executed depends on the operations it contains and the available hardware. Below you can see the speedup</div><div class="line"><a name="l00027"></a><span class="lineno"> 27</span> we've been experiencing in our experiments with a few common networks.</div><div class="line"><a name="l00028"></a><span class="lineno"> 28</span> </div><div class="line"><a name="l00029"></a><span class="lineno"> 29</span> <img align="center" width="700" src="https://developer.arm.com/-/media/developer/Other Images/Arm_NN_performance_relative_to_other_NN_frameworks_diagram.png"/></div><div class="line"><a name="l00030"></a><span class="lineno"> 30</span> </div><div class="line"><a name="l00031"></a><span class="lineno"> 31</span> Arm NN is written using portable C++14 and the build system uses [CMake](https://cmake.org/), therefore it is possible</div><div class="line"><a name="l00032"></a><span class="lineno"> 32</span> to build for a wide variety of target platforms, from a wide variety of host environments.</div><div class="line"><a name="l00033"></a><span class="lineno"> 33</span> </div><div class="line"><a name="l00034"></a><span class="lineno"> 34</span> </div><div class="line"><a name="l00035"></a><span class="lineno"> 35</span> ## Getting started: Software tools overview</div><div class="line"><a name="l00036"></a><span class="lineno"> 36</span> Depending on what kind of framework (Tensorflow, Caffe, ONNX) you've been using to create your model there are multiple</div><div class="line"><a name="l00037"></a><span class="lineno"> 37</span> software tools available within Arm NN that can serve your needs.</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> Generally, there is a **parser** available **for each supported framework**. Each parser allows you to run models from</div><div class="line"><a name="l00040"></a><span class="lineno"> 40</span> one framework e.g. the TfLite-Parser lets you run TfLite models. You can integrate these parsers into your own</div><div class="line"><a name="l00041"></a><span class="lineno"> 41</span> application to load, optimize and execute your model. We also provide **python bindings** for our parsers and the Arm NN core.</div><div class="line"><a name="l00042"></a><span class="lineno"> 42</span> We call the result **PyArmNN**. Therefore your application can be conveniently written in either C++ using the "original"</div><div class="line"><a name="l00043"></a><span class="lineno"> 43</span> Arm NN library or in Python using PyArmNN. You can find tutorials on how to setup and use our parsers in our doxygen</div><div class="line"><a name="l00044"></a><span class="lineno"> 44</span> documentation. The latest version can be found in the [wiki section](https://github.com/ARM-software/armnn/wiki/Documentation)</div><div class="line"><a name="l00045"></a><span class="lineno"> 45</span> of this repository.</div><div class="line"><a name="l00046"></a><span class="lineno"> 46</span> </div><div class="line"><a name="l00047"></a><span class="lineno"> 47</span> Admittedly, building Arm NN and its parsers from source is not always easy to accomplish. We are trying to increase our</div><div class="line"><a name="l00048"></a><span class="lineno"> 48</span> usability by providing Arm NN as a **Debian package**. Our debian package is the most easy way to install the Arm NN Core,</div><div class="line"><a name="l00049"></a><span class="lineno"> 49</span> the TfLite Parser and PyArmNN (More support is about to come): [Installation via Apt Repository](InstallationViaAptRepository.md)</div><div class="line"><a name="l00050"></a><span class="lineno"> 50</span> </div><div class="line"><a name="l00051"></a><span class="lineno"> 51</span> The newest member in Arm NNs software toolkit is the **TfLite Delegate**. The delegate can be integrated in TfLite.</div><div class="line"><a name="l00052"></a><span class="lineno"> 52</span> TfLite will then delegate operations, that can be accelerated with Arm NN, to Arm NN. Every other operation will still be</div><div class="line"><a name="l00053"></a><span class="lineno"> 53</span> executed with the usual TfLite runtime. This is our **recommended way to accelerate TfLite models**. As with our parsers</div><div class="line"><a name="l00054"></a><span class="lineno"> 54</span> there are tutorials in our doxygen documentation that can be found in the [wiki section](https://github.com/ARM-software/armnn/wiki/Documentation).</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> If you would like to use **Arm NN on Android** you can follow this guide which explains [how to build Arm NN using the AndroidNDK]().</div><div class="line"><a name="l00057"></a><span class="lineno"> 57</span> But you might also want to take a look at another repository which implements a hardware abstraction layer (HAL) for</div><div class="line"><a name="l00058"></a><span class="lineno"> 58</span> Android. The repository is called [Android-NN-Driver](https://github.com/ARM-software/android-nn-driver) and when</div><div class="line"><a name="l00059"></a><span class="lineno"> 59</span> integrated into Android it will automatically run neural networks with Arm NN.</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> </div><div class="line"><a name="l00062"></a><span class="lineno"> 62</span> ## Where to find more information</div><div class="line"><a name="l00063"></a><span class="lineno"> 63</span> The section above introduces the most important tools that Arm NN provides.</div><div class="line"><a name="l00064"></a><span class="lineno"> 64</span> You can find a complete list in our **doxygen documentation**. The</div><div class="line"><a name="l00065"></a><span class="lineno"> 65</span> latest version can be found in the [wiki section](https://github.com/ARM-software/armnn/wiki/Documentation) of our github</div><div class="line"><a name="l00066"></a><span class="lineno"> 66</span> repository.</div><div class="line"><a name="l00067"></a><span class="lineno"> 67</span> </div><div class="line"><a name="l00068"></a><span class="lineno"> 68</span> For FAQs and troubleshooting advice, see [FAQ.md](docs/FAQ.md)</div><div class="line"><a name="l00069"></a><span class="lineno"> 69</span> or take a look at previous [github issues](https://github.com/ARM-software/armnn/issues).</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> </div><div class="line"><a name="l00072"></a><span class="lineno"> 72</span> ## Note</div><div class="line"><a name="l00073"></a><span class="lineno"> 73</span> 1. The following tools are **deprecated** in Arm NN 21.02 and will be removed in 21.05:</div><div class="line"><a name="l00074"></a><span class="lineno"> 74</span>  * TensorflowParser</div><div class="line"><a name="l00075"></a><span class="lineno"> 75</span>  * CaffeParser</div><div class="line"><a name="l00076"></a><span class="lineno"> 76</span>  * Quantizer</div><div class="line"><a name="l00077"></a><span class="lineno"> 77</span> </div><div class="line"><a name="l00078"></a><span class="lineno"> 78</span> 2. Ubuntu Linux 16.04 LTS will no longer be supported by April 30, 2021.</div><div class="line"><a name="l00079"></a><span class="lineno"> 79</span>  At that time, Ubuntu 16.04 LTS will no longer receive security patches or other software updates.</div><div class="line"><a name="l00080"></a><span class="lineno"> 80</span>  Consequently Arm NN will from the 21.08 Release at the end of August 2021 no longer be officially</div><div class="line"><a name="l00081"></a><span class="lineno"> 81</span>  supported on Ubuntu 16.04 LTS but will instead be supported on Ubuntu 18.04 LTS.</div><div class="line"><a name="l00082"></a><span class="lineno"> 82</span> </div><div class="line"><a name="l00083"></a><span class="lineno"> 83</span> 3. We are currently in the process of removing [boost](https://www.boost.org/) as a dependency to Arm NN. This process</div><div class="line"><a name="l00084"></a><span class="lineno"> 84</span>  is finished for everything apart from our unit tests. This means you don't need boost to build and use Arm NN but</div><div class="line"><a name="l00085"></a><span class="lineno"> 85</span>  you need it to execute our unit tests. Boost will soon be removed from Arm NN entirely.</div><div class="line"><a name="l00086"></a><span class="lineno"> 86</span> </div><div class="line"><a name="l00087"></a><span class="lineno"> 87</span> </div><div class="line"><a name="l00088"></a><span class="lineno"> 88</span> ## Contributions</div><div class="line"><a name="l00089"></a><span class="lineno"> 89</span> The Arm NN project welcomes contributions. For more details on contributing to Arm NN see the [Contributing page](https://mlplatform.org/contributing/)</div><div class="line"><a name="l00090"></a><span class="lineno"> 90</span> on the [MLPlatform.org](https://mlplatform.org/) website, or see the [Contributor Guide](ContributorGuide.md).</div><div class="line"><a name="l00091"></a><span class="lineno"> 91</span> </div><div class="line"><a name="l00092"></a><span class="lineno"> 92</span> Particularly if you'd like to implement your own backend next to our CPU, GPU and NPU backends there are guides for</div><div class="line"><a name="l00093"></a><span class="lineno"> 93</span> backend development:</div><div class="line"><a name="l00094"></a><span class="lineno"> 94</span> [Backend development guide](src/backends/README.md), [Dynamic backend development guide](src/dynamic/README.md)</div><div class="line"><a name="l00095"></a><span class="lineno"> 95</span> </div><div class="line"><a name="l00096"></a><span class="lineno"> 96</span> </div><div class="line"><a name="l00097"></a><span class="lineno"> 97</span> ## Disclaimer</div><div class="line"><a name="l00098"></a><span class="lineno"> 98</span> The armnn/tests directory contains tests used during Arm NN development. Many of them depend on third-party IP, model</div><div class="line"><a name="l00099"></a><span class="lineno"> 99</span> protobufs and image files not distributed with Arm NN. The dependencies of some of the tests are available freely on</div><div class="line"><a name="l00100"></a><span class="lineno"> 100</span> the Internet, for those who wish to experiment, but they won't run out of the box.</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> </div><div class="line"><a name="l00103"></a><span class="lineno"> 103</span> ## License</div><div class="line"><a name="l00104"></a><span class="lineno"> 104</span> Arm NN is provided under the [MIT](https://spdx.org/licenses/MIT.html) license.</div><div class="line"><a name="l00105"></a><span class="lineno"> 105</span> See [LICENSE](LICENSE) for more information. Contributions to this project are accepted under the same license.</div><div class="line"><a name="l00106"></a><span class="lineno"> 106</span> </div><div class="line"><a name="l00107"></a><span class="lineno"> 107</span> Individual files contain the following tag instead of the full license text.</div><div class="line"><a name="l00108"></a><span class="lineno"> 108</span> </div><div class="line"><a name="l00109"></a><span class="lineno"> 109</span>  SPDX-License-Identifier: MIT</div><div class="line"><a name="l00110"></a><span class="lineno"> 110</span> </div><div class="line"><a name="l00111"></a><span class="lineno"> 111</span> This enables machine processing of license information based on the SPDX License Identifiers that are available here: http://spdx.org/licenses/</div><div class="line"><a name="l00112"></a><span class="lineno"> 112</span> </div><div class="line"><a name="l00113"></a><span class="lineno"> 113</span> </div><div class="line"><a name="l00114"></a><span class="lineno"> 114</span> ## Third-party</div><div class="line"><a name="l00115"></a><span class="lineno"> 115</span> Third party tools used by Arm NN:</div><div class="line"><a name="l00116"></a><span class="lineno"> 116</span> </div><div class="line"><a name="l00117"></a><span class="lineno"> 117</span> | Tool | License (SPDX ID) | Description | Version | Provenience</div><div class="line"><a name="l00118"></a><span class="lineno"> 118</span> |----------------|-------------------|------------------------------------------------------------------|-------------|-------------------</div><div class="line"><a name="l00119"></a><span class="lineno"> 119</span> | cxxopts | MIT | A lightweight C++ option parser library | SHA 12e496da3d486b87fa9df43edea65232ed852510 | https://github.com/jarro2783/cxxopts</div><div class="line"><a name="l00120"></a><span class="lineno"> 120</span> | fmt | MIT | {fmt} is an open-source formatting library providing a fast and safe alternative to C stdio and C++ iostreams. | 7.0.1 | https://github.com/fmtlib/fmt</div><div class="line"><a name="l00121"></a><span class="lineno"> 121</span> | ghc | MIT | A header-only single-file std::filesystem compatible helper library | 1.3.2 | https://github.com/gulrak/filesystem</div><div class="line"><a name="l00122"></a><span class="lineno"> 122</span> | half | MIT | IEEE 754 conformant 16-bit half-precision floating point library | 1.12.0 | http://half.sourceforge.net</div><div class="line"><a name="l00123"></a><span class="lineno"> 123</span> | mapbox/variant | BSD | A header-only alternative to 'boost::variant' | 1.1.3 | https://github.com/mapbox/variant</div><div class="line"><a name="l00124"></a><span class="lineno"> 124</span> | stb | MIT | Image loader, resize and writer | 2.16 | https://github.com/nothings/stb</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="_r_e_a_d_m_e_8md.xhtml">README.md</a></li> - <li class="footer">Generated on Thu Feb 25 2021 17:27:54 for ArmNN by + <li class="footer">Generated on Fri Mar 19 2021 15:26:05 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> |