Low-rank Gradient Approximation For Memory-Efficient On-device Training of Deep Neural Network

Open in new window