YU, Hao; YANG, Sen; ZHU, Shenghuo. Parallel Restarted SGD with Faster Convergence and Less Communication: Demystifying Why Model Averaging Works for Deep Learning. Proceedings of the AAAI Conference on Artificial Intelligence, [S. l.], v. 33, n. 01, p. 5693–5700, 2019. DOI: 10.1609/aaai.v33i01.33015693. Disponível em: https://ojs.aaai.org/index.php/AAAI/article/view/4514. Acesso em: 8 may. 2026.