Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2405.14669 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912099088728064 |
|---|---|
| author | Sun, Peng Jiang, Yi Lin, Tao |
| author_facet | Sun, Peng Jiang, Yi Lin, Tao |
| contents | 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}. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2405_14669 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Efficiency for Free: Ideal Data Are Transportable Representations Sun, Peng Jiang, Yi Lin, Tao Machine Learning Artificial Intelligence 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}. |
| title | Efficiency for Free: Ideal Data Are Transportable Representations |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2405.14669 |