Data-Parallel Neural Network Training via Nonlinearly Preconditioned Trust-Region Method

Open in new window