at Google ScholarTitle data
Isenko, Alexander ; Mayer, Ruben ; Jedele, Jeffrey ; Jacobsen, Hans-Arno:
Where Is My Training Bottleneck? Hidden Trade-Offs in Deep Learning Preprocessing Pipelines.
In:
Proceedings of the 2022 International Conference on Management of Data. -
New York
: Association for Computing Machinery
,
2022
. - pp. 1825-1839
ISBN 978-1-4503-9249-5
DOI: https://doi.org/10.1145/3514221.3517848
Abstract in another language
Preprocessing pipelines in deep learning aim to provide sufficient data throughput to keep the training processes busy. Maximizing resource utilization is becoming more challenging as the throughput of training processes increases with hardware innovations (e.g., faster GPUs, TPUs, and inter-connects) and advanced parallelization techniques that yield better scalability. At the same time, the amount of training data needed in order to train increasingly complex models is growing. As a consequence of this development, data preprocessing and provisioning are becoming a severe bottleneck in end-to-end deep learning pipelines.
In this paper, we provide an in-depth analysis of data preprocessing pipelines from four different machine learning domains. We introduce a new perspective on efficiently preparing datasets for end-to-end deep learning pipelines and extract individual trade-offs to optimize throughput, preprocessing time, and storage consumption. Additionally, we provide an open-source profiling library that can automatically decide on a suitable preprocessing strategy to maximize throughput. By applying our generated insights to real-world use-cases, we obtain an increased throughput of 3x to 13x compared to an untuned system while keeping the pipeline functionally identical. These findings show the enormous potential of data pipeline tuning.
Further data
| Item Type: | Article in a book |
|---|---|
| Refereed: | Yes |
| Institutions of the University: | Faculties Faculties > Faculty of Mathematics, Physics und Computer Science Faculties > Faculty of Mathematics, Physics und Computer Science > Department of Computer Science > Chair Data Systems Faculties > Faculty of Mathematics, Physics und Computer Science > Department of Computer Science > Chair Data Systems > Chair Data Systems - Univ.-Prof. Dr. Ruben Mayer Faculties > Faculty of Mathematics, Physics und Computer Science > Department of Computer Science |
| Result of work at the UBT: | No |
| DDC Subjects: | 000 Computer Science, information, general works > 004 Computer science |
| Date Deposited: | 26 Apr 2023 11:38 |
| Last Modified: | 05 Feb 2024 07:32 |
| URI: | https://eref.uni-bayreuth.de/id/eprint/76044 |