{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/training-a-large-video-model-on-a-single","title":"Training a Large Video Model on a Single Machine in a Day","arxiv_id":"2309.16669","date":"2023-09-28","proceeding":null,"authors":["Yue Zhao","Philipp Krähenbühl"],"abstract":"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.","url_abs":"https://arxiv.org/abs/2309.16669v1","url_pdf":"https://arxiv.org/pdf/2309.16669v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"training-a-large-video-model-on-a-single","repo_url":"https://github.com/zhaoyue-zephyrus/avion","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"action-recognition-in-videos","task_name":"Action Recognition"},{"task_slug":null,"task_name":"CPU"},{"task_slug":null,"task_name":"GPU"},{"task_slug":"multi-instance-retrieval","task_name":"Multi-Instance Retrieval"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/action-recognition-on-epic-kitchens-100","task":"Action Recognition","dataset":"EPIC-KITCHENS-100","model":"Avion (ViT-L)","rank_in_archive_order":3,"of":32,"metrics":{"Action@1":"54.4","Noun@1":"65.4","Verb@1":"73.0"},"uses_additional_data":true}],"syntology":{"syntology_url":"https://syntology.ai/paper/2309.16669","atlas_url":"https://app.syntology.ai/?focus=2309.16669","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}