{"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/net2net-accelerating-learning-via-knowledge","title":"Net2Net: Accelerating Learning via Knowledge Transfer","arxiv_id":"1511.05641","date":"2015-11-18","proceeding":null,"authors":["Tianqi Chen","Ian Goodfellow","Jonathon Shlens"],"abstract":"We introduce techniques for rapidly transferring the information stored in\none neural net into another neural net. The main purpose is to accelerate the\ntraining of a significantly larger neural net. During real-world workflows, one\noften trains very many different neural networks during the experimentation and\ndesign process. This is a wasteful process in which each new model is trained\nfrom scratch. Our Net2Net technique accelerates the experimentation process by\ninstantaneously transferring the knowledge from a previous network to each new\ndeeper or wider network. Our techniques are based on the concept of\nfunction-preserving transformations between neural network specifications. This\ndiffers from previous approaches to pre-training that altered the function\nrepresented by a neural net when adding layers to it. Using our knowledge\ntransfer mechanism to add depth to Inception modules, we demonstrate a new\nstate of the art accuracy rating on the ImageNet dataset.","url_abs":"http://arxiv.org/abs/1511.05641v4","url_pdf":"http://arxiv.org/pdf/1511.05641v4.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":"net2net-accelerating-learning-via-knowledge","repo_url":"https://github.com/agongt408/vbranch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"net2net-accelerating-learning-via-knowledge","repo_url":"https://github.com/hxtruong/net2net","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"net2net-accelerating-learning-via-knowledge","repo_url":"https://github.com/soumith/net2net.torch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"torch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"transfer-learning","task_name":"Transfer Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1511.05641","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}