This is a fully self-contained example using the NNoM library located here: https://github.com/majianjia/nnom
- CMake
- Required for building the C implementation of the NNOM model
- Python
- numpy
- tensorflow
- keras
- sklearn
- Start by building a model in python. Run
./tools/RunPython.ps1
to train the model and output the necessary files for the C implementation. - Run
./tools/BuildC.ps1
to build the C implementation of the model. This will output a .exe file. - Run
./tools/RunC.ps1
to run the C implementation of the model. This will output the results of the model. - Run
./tools/RunCompare.ps1
to compare the results of the python and C implementations. This will output the results of the comparison.
Note: The previous four steps can be done with one command ./tools/Pipeline.ps1
. All output will be saved to a file called output.log
in the root directory.