.. testing-libngraph: ########################## Testing the nGraph library ########################## The |InG| library code base uses the `GTest framework`_ for unit tests. CMake automatically downloads a copy of the required GTest files when configuring the build directory. To perform the unit tests: #. Create and configure the build directory as described in our :doc:`installation` guide. #. Enter the build directory and run ``make check``: .. code-block:: console $ cd build/ $ make check Compiling a framework with ``libngraph`` ======================================== After building and installing the nGraph library to your system, the next logical step is to compile a framework that you can use to run a training/inference model with one of the backends that are now enabled. For this early release, we provide integration guides for * `MXNet`_, * `TensorFlow`_, and * neon™ `frontend framework`_ Integration guides for each of these other frameworks is tentatively forthcoming and/or open to the community for contributions and sample documentation: * `Chainer`_, * `PyTorch`_, * `Caffe2`_, and * Frameworks not yet written (for algorithms that do not yet exist). .. _GTest framework: https://github.com/google/googletest.git .. _MXNet: http://mxnet.incubator.apache.org/ .. _TensorFlow: https://www.tensorflow.org/ .. _Caffe2: https://github.com/caffe2/ .. _PyTorch: http://pytorch.org/ .. _Chainer: https://chainer.org/ .. _frontend framework: http://neon.nervanasys.com/index.html/