1 paper
Haoli Bai, Jiaxiang Wu, Irwin King +1
Model compression has been widely adopted to obtain light-weighted deep neural networks. Most prevalent methods, however, require fine-tuning with sufficient training data to ensur…