{"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/simple-and-efficient-architecture-search-for","title":"Simple And Efficient Architecture Search for Convolutional Neural Networks","arxiv_id":"1711.04528","date":"2017-11-13","proceeding":"ICLR 2018 1","authors":["Thomas Elsken","Jan-Hendrik Metzen","Frank Hutter"],"abstract":"Neural networks have recently had a lot of success for many tasks. However,\nneural network architectures that perform well are still typically designed\nmanually by experts in a cumbersome trial-and-error process. We propose a new\nmethod to automatically search for well-performing CNN architectures based on a\nsimple hill climbing procedure whose operators apply network morphisms,\nfollowed by short optimization runs by cosine annealing. Surprisingly, this\nsimple method yields competitive results, despite only requiring resources in\nthe same order of magnitude as training a single network. E.g., on CIFAR-10,\nour method designs and trains networks with an error rate below 6% in only 12\nhours on a single GPU; training for one day reduces this error further, to\nalmost 5%.","url_abs":"http://arxiv.org/abs/1711.04528v1","url_pdf":"http://arxiv.org/pdf/1711.04528v1.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":"simple-and-efficient-architecture-search-for","repo_url":"https://github.com/famishedrover/NAS","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"simple-and-efficient-architecture-search-for","repo_url":"https://github.com/zhengjian2322/net2net","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":null,"task_name":"GPU"},{"task_slug":"architecture-search","task_name":"Neural Architecture Search"}],"methods":[{"method_slug":"neural-architecture-search","method_name":"Neural Architecture Search"},{"method_slug":"softmax","method_name":"Softmax"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1711.04528","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}