Weightless Neural Networks Library

• A Collecation of Weghtless models implementd in in paralellized C++ code with a python binding.
  • WiSARD
  • Bleaching WiSARD
  • Bloom WiSARD
  • Bleaching Bloom WiSARD
  • PseudoConvWiSARD
  • etc
• A Collectioins of Datasets commonly used in Weightless models
• and many more.

Compilation:

cd wzero/models/wrappers
cmake CMakeLists.txt
make

Mnist Example:

import wzero as wz

dataset = wz.datasets.Mnist().distrib_thermometer(10).flatten()
model = wz.models.WiSARD2(7840, 40, 10)
data = wz.experiments2.profile(dataset, model)

print(data)