Efficiency for Free: Ideal Data Are Transportable Representations
This work addresses the problem of high computational costs in machine learning for researchers and practitioners, offering a method to accelerate representation learning with significant efficiency gains.
The paper tackles the challenge of data efficiency in representation learning by showing that using a publicly available, task-agnostic prior model can generate efficient data, reducing computational costs by 50% on ImageNet-1K while maintaining accuracy compared to baseline methods.
Data, the seminal opportunity and challenge in modern machine learning, currently constrains the scalability of representation learning and impedes the pace of model evolution. In this work, we investigate the efficiency properties of data from both optimization and generalization perspectives. Our theoretical and empirical analysis reveals an unexpected finding: for a given task, utilizing a publicly available, task- and architecture-agnostic model (referred to as the `prior model' in this paper) can effectively produce efficient data. Building on this insight, we propose the Representation Learning Accelerator (\algopt), which promotes the formation and utilization of efficient data, thereby accelerating representation learning. Utilizing a ResNet-18 pre-trained on CIFAR-10 as a prior model to inform ResNet-50 training on ImageNet-1K reduces computational costs by 50% while maintaining the same accuracy as the model trained with the original BYOL, which requires 100% cost. Our code is available at: \url{https://github.com/LINs-lab/ReLA}.