{"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/you-only-search-once-single-shot-neural","title":"You Only Search Once: Single Shot Neural Architecture Search via Direct Sparse Optimization","arxiv_id":"1811.01567","date":"2018-11-05","proceeding":null,"authors":["Xinbang Zhang","Zehao Huang","Naiyan Wang"],"abstract":"Recently Neural Architecture Search (NAS) has aroused great interest in both\nacademia and industry, however it remains challenging because of its huge and\nnon-continuous search space. Instead of applying evolutionary algorithm or\nreinforcement learning as previous works, this paper proposes a Direct Sparse\nOptimization NAS (DSO-NAS) method. In DSO-NAS, we provide a novel model pruning\nview to NAS problem. In specific, we start from a completely connected block,\nand then introduce scaling factors to scale the information flow between\noperations. Next, we impose sparse regularizations to prune useless connections\nin the architecture. Lastly, we derive an efficient and theoretically sound\noptimization method to solve it. Our method enjoys both advantages of\ndifferentiability and efficiency, therefore can be directly applied to large\ndatasets like ImageNet. Particularly, On CIFAR-10 dataset, DSO-NAS achieves an\naverage test error 2.84\\%, while on the ImageNet dataset DSO-NAS achieves\n25.4\\% test error under 600M FLOPs with 8 GPUs in 18 hours.","url_abs":"http://arxiv.org/abs/1811.01567v1","url_pdf":"http://arxiv.org/pdf/1811.01567v1.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":"you-only-search-once-single-shot-neural","repo_url":"https://github.com/XinbangZhang/DSO-NAS","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"mxnet","reach":{"status":"ok"}}],"tasks":[{"task_slug":"architecture-search","task_name":"Neural Architecture Search"},{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"}],"methods":[{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1811.01567","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}