With the rapid development of storage and computing power on mobile devices, it becomes critical and popular to deploy models on devices to save onerous communication latencies and to capture real-time features. While quite a lot of works have explored to facilitate on-device learning and inference, most of them focus on dealing with response delay or privacy protection. Little has been done to model the collaboration between the device and the cloud modeling and benefit both sides jointly. To bridge this gap, we are among the first attempts to study the Device-Cloud Collaborative Learning (DCCL) framework. Specifically, we propose a novel MetaPatch learning approach on the device side to efficiently achieve "thousands of people with thousands of models" given a centralized cloud model. Then, with billions of updated personalized device models, we propose a "model-over-models" distillation algorithm, namely MoMoDistill, to update the centralized cloud model. Our extensive experiments over a range of datasets with different settings demonstrate the effectiveness of such collaboration on both cloud and devices, especially its superiority to model long-tailed users.
翻译:随着移动设备存储和计算能力的迅速发展,在设备上部署模型以节省繁琐的通信迟滞和捕捉实时功能变得至关重要和流行。虽然大量工作已经探索了便利在设备上学习和推断,但大部分侧重于应对反应延迟或隐私保护。随着设备与云型建模之间的协作模式的迅速发展,我们几乎没有做任何建模工作,因此双方都能共同受益。为了缩小这一差距,我们是第一批尝试研究设备用户协作学习框架(CDCL)的尝试之一。具体地说,我们提议在设备方面采用新的Metapatch学习方法,以高效地实现“数千个有数千个模型的人 ”, 以集中的云型模型为模式。随后,我们用数十亿个更新的个人化设备模型,提出了“模型” 蒸馏算法,即Modistill,以更新中央云型模型。我们对不同环境的一系列数据集进行的广泛实验,显示了在云和装置上的合作的有效性,特别是其优于模型的长期用户。