Training a Large Video Model on a Single Machine in a Day

Autor:	Zhao, Yue, Krähenbühl, Philipp
Rok vydání:	2023
Předmět:	Computer Science - Computer Vision and Pattern Recognition
Druh dokumentu:	Working Paper
Popis:	Videos are big, complex to pre-process, and slow to train on. State-of-the-art large-scale video models are trained on clusters of 32 or more GPUs for several days. As a consequence, academia largely ceded the training of large video models to industry. In this paper, we show how to still train a state-of-the-art video model on a single machine with eight consumer-grade GPUs in a day. We identify three bottlenecks, IO, CPU, and GPU computation, and optimize each. The result is a highly efficient video training pipeline. For comparable architectures, our pipeline achieves higher accuracies with $\frac{1}{8}$ of the computation compared to prior work. Code is available at https://github.com/zhaoyue-zephyrus/AVION. Comment: Tech report. Code is available at https://github.com/zhaoyue-zephyrus/AVION
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2309.16669 Zobrazit plný text záznamu View this record from Arxiv