NNTrainer 0.3.0 Release
We are releasing NNTrainer v0.3.0
RPM Files are for Tizen, built daily at build.tizen.org (https://build.tizen.org/package/show/Tizen:Unified/nntrainer)
and available at download.tizen.org. ( http://download.tizen.org/snapshots/tizen/unified/latest/repos/standard/packages/ )
DEB files are for Ubuntu, built and download from launchpad.net (https:://launchpad.net/~nnstreamer/+archive/ubuntu/ppa )
If you have unresoved dependencies, please download them from Ubuntu universe and nnstreamer PPA
In this release:
Fixes
- Fix Batch Normalization Bugs
- Fix Stride and Padding in Conv2D and Pooling2D Layer
- Add a lot of unit tests to evaluate NNTrainer implementation
and more.
New Features
- New Layers
- Recurrent Layers : RNN, LSTM, GRU
- Embedding Layer
- Distributed Layer
- KNN Layer
- L2Norm Layer
- Rewrite
DataSet
to support element-wise ( not batch-wise ) getter & Better Data Handling - Interpreter to convert the model into various other framework such as TfLite
- Provides More C/C++ APIs
- Inferece APIs
- Save / Load Model APIs
- New Applications
and more