Accelerating Hessian-free optimization for deep neural networks by implicit preconditioning and sampling

Open in new window