QGOpt: Riemannian optimization for quantum technologies

Many theoretical problems in quantum technology can be formulated and addressed as constrained optimization problems. The most common mechanical constraints such as, e.g., orthogonality of isometric unitary matrices, CPTP property channels, conditions on density seen quotient or embedded Riemannian manifolds. This allows to use techniques for solving quantum-mechanical In the present work, we introduce QGOpt, library technology. QGOpt relies underlying structure permits application standard gradient based methods while preserving constraints. Moreover, is written top TensorFlow, which enables automatic differentiation calculate necessary gradients optimization. We show two examples: gate decomposition tomography.