VAMO: Efficient Large-Scale Nonconvex Optimization via Adaptive Zeroth Order Variance ReductionShare on Twitter Facebook LinkedIn Previous Next