{"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/model-compression-as-constrained-optimization-1","title":"Model compression as constrained optimization, with application to neural nets. Part I: general framework","arxiv_id":"1707.01209","date":"2017-07-05","proceeding":null,"authors":["Miguel Á. Carreira-Perpiñán"],"abstract":"Compressing neural nets is an active research problem, given the large size\nof state-of-the-art nets for tasks such as object recognition, and the\ncomputational limits imposed by mobile devices. We give a general formulation\nof model compression as constrained optimization. This includes many types of\ncompression: quantization, low-rank decomposition, pruning, lossless\ncompression and others. Then, we give a general algorithm to optimize this\nnonconvex problem based on the augmented Lagrangian and alternating\noptimization. This results in a \"learning-compression\" algorithm, which\nalternates a learning step of the uncompressed model, independent of the\ncompression type, with a compression step of the model parameters, independent\nof the learning task. This simple, efficient algorithm is guaranteed to find\nthe best compressed model for the task in a local sense under standard\nassumptions.\n  We present separately in several companion papers the development of this\ngeneral framework into specific algorithms for model compression based on\nquantization, pruning and other variations, including experimental results on\ncompressing neural nets and other models.","url_abs":"http://arxiv.org/abs/1707.01209v1","url_pdf":"http://arxiv.org/pdf/1707.01209v1.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":"model-compression-as-constrained-optimization-1","repo_url":"https://github.com/UCMerced-ML/LC-model-compression","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"model-compression","task_name":"Model Compression"},{"task_slug":"object-recognition","task_name":"Object Recognition"},{"task_slug":"quantization","task_name":"Quantization"}],"methods":[{"method_slug":"pruning","method_name":"Pruning"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1707.01209","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}