Neural architecture search (NAS) has demonstrated amazing success in searching for efficient deep neural networks (DNNs) from a given supernet. In parallel, the lottery ticket hypothesis has shown that DNNs contain small subnetworks that can be trained from scratch to achieve a comparable or higher accuracy than original DNNs. As such, it is currently a common practice to develop efficient DNNs via a pipeline of first search and then prune. Nevertheless, doing so often requires a search-train-prune-retrain process and thus prohibitive computational cost. In this paper, we discover for the first time that both efficient DNNs and their lottery subnetworks (i.e., lottery tickets) can be directly identified from a supernet, which we term as SuperTickets, via a two-in-one training scheme with jointly architecture searching and parameter pruning. Moreover, we develop a progressive and unified SuperTickets identification strategy that allows the connectivity of subnetworks to change during supernet training, achieving better accuracy and efficiency trade-offs than conventional sparse training. Finally, we evaluate whether such identified SuperTickets drawn from one task can transfer well to other tasks, validating their potential of handling multiple tasks simultaneously. Extensive experiments and ablation studies on three tasks and four benchmark datasets validate that our proposed SuperTickets achieve boosted accuracy and efficiency trade-offs than both typical NAS and pruning pipelines, regardless of having retraining or not. Codes and pretrained models are available at https://github.com/RICE-EIC/SuperTickets.
翻译:神经架构搜索(NAS)在从某个超级网络寻找高效的深神经网络(DNN)方面表现出惊人的成功。 与此同时,彩票假设表明,DNN包含小型子网络,可以从零开始接受培训,以达到与原DNN的相似或更高精度。因此,目前的做法是通过先搜索再运行管道来开发高效的DNN。然而,这样做往往需要一个搜索-train-prun-retrain 进程,从而导致高昂的计算成本。在本文中,我们第一次发现,高效的DNN和他们的彩票子网络(即彩票)都可以直接从超级网络中识别出,从头到头,从头到尾,从头到尾,我们称为超级Ticket的小型小网络。因此,目前的做法是通过一个双对一的培训计划,联合搜索和参数运行。此外,我们开发了一个渐进和统一的超级Ticket识别战略,允许子网络在超级网络培训期间进行互连通,实现更好的准确性和效率交易。最后,我们评估的是,SUPTickets在一项任务上是否能够同时提升S-reallicreal exal lavelildal Stal Stal Stal Stal Stal 和另外一项任务。