Module kmath-noa

This module provides a kotlin-jvm frontend for the NOA library together with relevant functionality from LibTorch.

Our aim is to create a Bayesian computational platform which covers a wide set of applications from particle physics simulations to deep learning and general differentiable programs written on top of AutoGrad & ATen.

Currently, the native artifacts support only GNU and CUDA for GPU acceleration.