{"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/scalable-gradient-based-tuning-of-continuous","title":"Scalable Gradient-Based Tuning of Continuous Regularization Hyperparameters","arxiv_id":"1511.06727","date":"2015-11-20","proceeding":null,"authors":["Jelena Luketina","Mathias Berglund","Klaus Greff","Tapani Raiko"],"abstract":"Hyperparameter selection generally relies on running multiple full training\ntrials, with selection based on validation set performance. We propose a\ngradient-based approach for locally adjusting hyperparameters during training\nof the model. Hyperparameters are adjusted so as to make the model parameter\ngradients, and hence updates, more advantageous for the validation cost. We\nexplore the approach for tuning regularization hyperparameters and find that in\nexperiments on MNIST, SVHN and CIFAR-10, the resulting regularization levels\nare within the optimal regions. The additional computational cost depends on\nhow frequently the hyperparameters are trained, but the tested scheme adds only\n30% computational overhead regardless of the model size. Since the method is\nsignificantly less computationally demanding compared to similar gradient-based\napproaches to hyperparameter optimization, and consistently finds good\nhyperparameter values, it can be a useful tool for training neural network\nmodels.","url_abs":"http://arxiv.org/abs/1511.06727v3","url_pdf":"http://arxiv.org/pdf/1511.06727v3.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":"scalable-gradient-based-tuning-of-continuous","repo_url":"https://github.com/jelennal/t1t2","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"hyperparameter-optimization","task_name":"Hyperparameter Optimization"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1511.06727","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}