README.md

   1 # NNtrainer
   2
   3 [![Code Coverage](http://nnstreamer.mooo.com/nntrainer/ci/badge/codecoverage.svg)](http://nnstreamer.mooo.com/nntrainer/ci/gcov_html/index.html)
   4 ![GitHub repo size](https://img.shields.io/github/repo-size/nnstreamer/nntrainer)
   5 ![GitHub issues](https://img.shields.io/github/issues/nnstreamer/nntrainer)
   6 ![GitHub pull requests](https://img.shields.io/github/issues-pr/nnstreamer/nntrainer)
   7 <a href="https://scan.coverity.com/projects/nnstreamer-nntrainer">
   8   <img alt="Coverity Scan Build Status"
   9        src="https://scan.coverity.com/projects/22512/badge.svg"/>
  10 </a>
  11 [![DailyBuild](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/badge/daily_build_test_result_badge.svg)](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/build_result/)
  12
  13 NNtrainer is a Software Framework for training Neural Network models on devices.
  14
  15 ## Overview
  16
  17 NNtrainer is an Open Source Project. The aim of the NNtrainer is to develop a Software Framework to train neural network models on embedded devices which have relatively limited resources. Rather than training whole layers of a network from the scratch, NNtrainer finetunes the neural network model on device with user data for the personalization.
  18
  19 Even if NNtariner runs on device, it provides full functionalities to train models and also utilizes limited device resources efficiently. NNTrainer is able to train various machine learning algorithms such as k-Nearest Neighbor (k-NN), Neural Networks, Logistic Regression, Reinforcement Learning algorithms, Recurrent network and more. We also provide examples for various tasks such as Few-shot learning, ResNet, VGG, Product Rating and more will be added. All of these were tested on Samsung Galaxy smart phone with Android and PC (Ubuntu 18.04/20.04).
  20
  21 [ NNTrainer: Towards the on-device learning for personalization ](https://www.youtube.com/watch?v=HWiV7WbIM3E), Samsung Software Developer Conference 2021 (Korean) <br />
  22 [ NNTrainer: Personalize neural networks on devices! ](https://www.youtube.com/watch?v=HKKowY78P1A), Samsung Developer Conference 2021 <br />
  23 [ NNTrainer: "On-device learning" ](https://www.youtube.com/embed/Jy_auavraKg?start=4035&end=4080), Samsung AI Forum 2021
  24
  25 ## Official Releases
  26
  27 |     | [Tizen](http://download.tizen.org/snapshots/tizen/unified/latest/repos/standard/packages/) | [Ubuntu](https://launchpad.net/~nnstreamer/+archive/ubuntu/ppa) | Android/NDK Build |
  28 | :-- | :--: | :--: | :--: |
  29 |     | 6.0M2 and later | 18.04 | 9/P |
  30 | arm | [![armv7l badge](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/badge/tizen.armv7l_result_badge.svg)](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/build_result/) | Available  | Ready |
  31 | arm64 |  [![aarch64 badge](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/badge/tizen.aarch64_result_badge.svg)](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/build_result/) | Available  | [![android badge](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/badge/arm64_v8a_android_result_badge.svg)](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/build_result/) |
  32 | x64 | [![x64 badge](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/badge/tizen.x86_64_result_badge.svg)](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/build_result/)  | [![ubuntu badge](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/badge/ubuntu_result_badge.svg)](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/build_result/)  | Ready  |
  33 | x86 | [![x86 badge](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/badge/tizen.i586_result_badge.svg)](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/build_result/)  | N/A  | N/A  |
  34 | Publish | [Tizen Repo](http://download.tizen.org/snapshots/tizen/unified/latest/repos/standard/packages/) | [PPA](https://launchpad.net/~nnstreamer/+archive/ubuntu/ppa) |   |
  35 | API | C (Official) | C/C++ | C/C++  |
  36
  37 - Ready: CI system ensures build-ability and unit-testing. Users may easily build and execute. However, we do not have automated release & deployment system for this instance.
  38 - Available: binary packages are released and deployed automatically and periodically along with CI tests.
  39 - [Daily Release](http://nnstreamer.mooo.com/nntrainer/ci/daily-build/build_result/)
  40 - SDK Support: Tizen Studio (6.0 M2+)
  41
  42 ## Maintainer
  43 * [Jijoong Moon](https://github.com/jijoongmoon)
  44 * [MyungJoo Ham](https://github.com/myungjoo)
  45 * [Geunsik Lim](https://github.com/leemgs)
  46
  47 ## Reviewers
  48 * [Sangjung Woo](https://github.com/again4you)
  49 * [Wook Song](https://github.com/wooksong)
  50 * [Jaeyun Jung](https://github.com/jaeyun-jung)
  51 * [Hyoungjoo Ahn](https://github.com/helloahn)
  52 * [Parichay Kapoor](https://github.com/kparichay)
  53 * [Dongju Chae](https://github.com/dongju-chae)
  54 * [Gichan Jang](https://github.com/gichan-jang)
  55 * [Yongjoo Ahn](https://github.com/anyj0527)
  56 * [Jihoon Lee](https://github.com/zhoonit)
  57 * [Hyeonseok Lee](https://github.com/lhs8928)
  58 * [Mete Ozay](https://github.com/meteozay)
  59
  60 ## Components
  61
  62 ### Supported Layers
  63
  64 This component defines layers which consist of a neural network model. Layers have their own properties to be set.
  65
  66  | Keyword | Layer Class Name | Description |
  67  |:-------:|:---:|:---|
  68  | conv1d | Conv1DLayer | Convolution 1-Dimentional Layer |
  69  | conv2d | Conv2DLayer |Convolution 2-Dimentional Layer |
  70  | pooling2d | Pooling2DLayer |Pooling 2-Dimentional Layer. Support average / max / global average / global max pooling |
  71  | flatten | FlattenLayer | Flatten layer |
  72  | fully_connected | FullyConnectedLayer | Fully connected layer |
  73  | input | InputLayer | Input Layer.  This is not always required. |
  74  | batch_normalization | BatchNormalizationLayer | Batch normalization layer |
  75  | activation | ActivaitonLayer | Set by layer property |
  76  | addition | AdditionLayer | Add input input layers |
  77  | attention | AttentionLayer | Attenstion layer |
  78  | centroid_knn | CentroidKNN | Centroid K-nearest neighbor layer |
  79  | concat | ConcatLayer | Concatenate input layers |
  80  | multiout | MultiOutLayer | Multi-Output Layer |
  81  | backbone_nnstreamer | NNStreamerLayer | Encapsulate NNStreamer layer |
  82  | backbone_tflite | TfLiteLayer | Encapsulate tflite as an layer |
  83  | permute | PermuteLayer | Permute layer for transpose |
  84  | preprocess_flip | PreprocessFlipLayer | Preprocess random flip layer |
  85  | preprocess_l2norm | PreprocessL2NormLayer | Preprocess simple l2norm layer to normalize |
  86  | preprocess_translate | PreprocessTranslateLayer | Preprocess translate layer |
  87  | reshape | ReshapeLayer | Reshape tensor dimension layer |
  88  | split | SplitLayer | Split layer |
  89  | dropout | DropOutLayer | Dropout Layer |
  90  | embedding | EmbeddingLayer | Embedding Layer |
  91  | rnn | RNNLayer | Recurrent Layer |
  92  | gru | GRULayer | Gated Recurrent Unit Layer |
  93  | lstm | LSTMLayer | Long Short-Term Memory Layer |
  94  | lstmcell | LSTMCellLayer | Long Short-Term Memory Cell Layer |
  95  | time_dist | TimeDistLayer | Time distributed Layer |
  96
  97 ### Supported Optimizers
  98
  99 NNTrainer Provides
 100
 101  | Keyword | Optimizer Name | Description |
 102  |:-------:|:---:|:---:|
 103  | sgd | Stochastic Gradient Decent | - |
 104  | adam | Adaptive Moment Estimation | - |
 105
 106 ### Supported Loss Functions
 107
 108 NNTrainer provides
 109
 110  | Keyword | Class Name | Description |
 111  |:-------:|:---:|:---:|
 112  | cross_sigmoid | CrossEntropySigmoidLossLayer | Cross entropy sigmoid loss layer |
 113  | cross_softmax | CrossEntropySoftmaxLossLayer | Cross entropy softmax loss layer |
 114  | constant_derivative | ConstantDerivativeLossLayer | Constant derivative loss layer |
 115  | mse | MSELossLayer | Mean square error loss layer |
 116
 117 ### Supported Activation Functions
 118
 119 NNTrainer provides
 120
 121  | Keyword | Loss Name | Description |
 122  |:-------:|:---:|:---|
 123  | tanh | tanh function | set as layer property |
 124  | sigmoid | sigmoid function | set as layer property |
 125  | relu | relu function | set as layer propery |
 126  | softmax | softmax function | set as layer propery |
 127  | weight_initializer | Weight Initialization | Xavier(Normal/Uniform), LeCun(Normal/Uniform),  HE(Normal/Unifor) |
 128  | weight_regularizer | weight decay ( L2Norm only ) | needs set weight_regularizer_param & type |
 129  | learnig_rate_decay | learning rate decay | need to set step |
 130
 131 ### Tensor
 132
 133 Tensor is responsible for calculation of a layer. It executes several operations such as addition, division, multiplication, dot production, data averaging and so on. In order to accelerate  calculation speed, CBLAS (C-Basic Linear Algebra: CPU) and CUBLAS (CUDA: Basic Linear Algebra) for PC (Especially NVIDIA GPU) are implemented for some of the operations. Later, these calculations will be optimized.
 134 Currently, we supports lazy calculation mode to reduce complexity for copying tensors during calculations.
 135
 136  | Keyword | Description |
 137  |:-------:|:---:|
 138  | 4D Tensor | B, C, H, W|
 139  | Add/sub/mul/div | - |
 140  | sum, average, argmax | - |
 141  | Dot, Transpose | - |
 142  | normalization, standardization | - |
 143  | save, read | - |
 144
 145 ### Others
 146
 147 NNTrainer provides
 148
 149  | Keyword | Loss Name | Description |
 150  |:-------:|:---:|:---|
 151  | weight_initializer | Weight Initialization | Xavier(Normal/Uniform), LeCun(Normal/Uniform),  HE(Normal/Unifor) |
 152  | weight_regularizer | weight decay ( L2Norm only ) | needs set weight_regularizer_constant & type |
 153  | learnig_rate_decay | learning rate decay | need to set step |
 154
 155 ### APIs
 156 Currently, we provide [C APIs](https://github.com/nnstreamer/nntrainer/blob/master/api/capi/include/nntrainer.h) for Tizen. [C++ APIs](https://github.com/nnstreamer/nntrainer/blob/master/api/ccapi/include) are also provided for other platform. Java & C# APIs will be provided soon.
 157
 158
 159 ### Examples for NNTrainer
 160
 161 #### [Custom Shortcut Application](https://github.com/nnstreamer/nntrainer/tree/main/Applications/Tizen_native/CustomShortcut)
 162
 163
 164 A demo application which enable user defined custom shortcut on galaxy watch.
 165
 166 #### [MNIST Example](https://github.com/nnstreamer/nntrainer/tree/main/Applications/MNIST)
 167
 168 An example to train mnist dataset. It consists two convolution 2d layer, 2 pooling 2d layer, flatten layer and fully connected layer.
 169
 170 #### [Reinforcement Learning Example](https://github.com/nnstreamer/nntrainer/tree/main/Applications/ReinforcementLearning/DeepQ)
 171
 172 A reinforcement learning example with cartpole game. It is using DeepQ algorithm.
 173
 174 #### [Transfer Learning Examples](https://github.com/nnstreamer/nntrainer/tree/main/Applications/TransferLearning)
 175
 176 Transfer learning examples with for image classification using the Cifar 10 dataset and for OCR. TFlite is used for feature extractor and modify last layer (fully connected layer) of network.
 177
 178 #### [ResNet Example](https://github.com/nnstreamer/nntrainer/tree/main/Applications/Resnet)
 179
 180 An example to train resnet18 network.
 181
 182 #### [VGG Example](https://github.com/nnstreamer/nntrainer/tree/main/Applications/VGG)
 183
 184 An example to train vgg16 network.
 185
 186 #### [ProductRating Example](https://github.com/nnstreamer/nntrainer/tree/main/Applications/ProductRatings)
 187
 188 This application contains a simple embedding-based model that predicts ratings given a user and a product.
 189
 190 #### [SimpleShot Example](https://github.com/nnstreamer/nntrainer/tree/main/Applications/SimpleShot)
 191
 192 An example to demonstrate few-shot learning : SimpleShot
 193
 194 #### [Custom Example](https://github.com/nnstreamer/nntrainer/tree/main/Applications/Custom)
 195
 196 An example to demonstrate how to create custom layers, optimizers or other supported objects.
 197
 198 <!-- #### Tizen CAPI Example -->
 199
 200 <!-- An example to demonstrate c api for Tizen. It is same transfer learing but written with tizen c api.~ -->
 201 <!-- Deleted instead moved to a [test](https://github.com/nnstreamer/nntrainer/blob/master/test/tizen_capi/unittest_tizen_capi.cpp) -->
 202
 203 #### [KNN Example](https://github.com/nnstreamer/nntrainer/tree/main/Applications/KNN)
 204
 205 A transfer learning example with for image classification using the Cifar 10 dataset. TFlite is used for feature extractor and compared with KNN.
 206
 207 #### [Logistic Regression Example](https://github.com/nnstreamer/nntrainer/tree/main/Applications/LogisticRegression)
 208
 209 A logistic regression example using NNTrainer.
 210
 211 ## [Getting Started](https://github.com/nnstreamer/nntrainer/blob/main/docs/getting-started.md)
 212
 213 Instructions for installing NNTrainer.
 214
 215 ### [Running Examples](https://github.com/nnstreamer/nntrainer/blob/main/docs/how-to-run-examples.md)
 216
 217 Instructions for preparing NNTrainer for execution
 218
 219 ## Open Source License
 220
 221 The nntrainer is an open source project released under the terms of the Apache License version 2.0.
 222
 223 ## Contributing
 224
 225 Contributions are welcome! Please see our [Contributing](https://github.com/nnstreamer/nntrainer/blob/main/docs/contributing.md) Guide for more details.
 226
 227 [![](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/images/0)](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/links/0)[![](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/images/1)](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/links/1)[![](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/images/2)](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/links/2)[![](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/images/3)](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/links/3)[![](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/images/4)](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/links/4)[![](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/images/5)](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/links/5)[![](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/images/6)](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/links/6)[![](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/images/7)](https://sourcerer.io/fame/dongju-chae/nnstreamer/nntrainer/links/7)