{"id":3751,"date":"2025-04-24T14:49:15","date_gmt":"2025-04-24T19:49:15","guid":{"rendered":"https:\/\/wordpress.cels.anl.gov\/lans-seminars\/?post_type=tribe_events&#038;p=3751"},"modified":"2025-04-25T11:50:52","modified_gmt":"2025-04-25T16:50:52","slug":"3751","status":"publish","type":"tribe_events","link":"https:\/\/wordpress.cels.anl.gov\/lans-seminars\/event\/3751\/","title":{"rendered":"LANS Seminar"},"content":{"rendered":"<p><strong>Seminar Title: <\/strong>Improving the Worst-Case Bidirectional Communication Complexity for Nonconvex Distributed Optimization under Function Similarity<\/p>\n<p><strong>Speaker: <\/strong>Kaja Gruntkowska, PhD Student, Optimization, Machine Learning, KAUST<\/p>\n<p><strong>Date\/Time:<\/strong> Thursday, May 1, 2025\/ 11:00AM \u2013 12:00 PM (Virtual)<br \/>\n<strong>Location:\u00a0<\/strong><em>See Meeting URL on the cels-seminars website which will require an Argonne login.<\/em><\/p>\n<p><strong>Host: <\/strong>Krishnan Raghavan<\/p>\n<p class=\"p1\"><strong>Description: <\/strong>Effective communication between the server and workers plays a key role in distributed optimization. In this paper, we focus on optimizing the server-to-worker communication, uncovering inefficiencies in prevalent downlink compression approaches. Considering first the pure setup where the uplink communication costs are negligible, we introduce MARINA-P, a novel method for downlink compression, employing a collection of correlated compressors. Theoretical analysis demonstrates that MARINA-P with permutation compressors can achieve a server-to-worker communication complexity improving with the number of workers, thus being provably superior to existing algorithms. We further show that MARINA-P can serve as a starting point for extensions such as methods supporting bidirectional compression. We introduce M3, a method combining MARINA-P with uplink compression and a momentum step, achieving bidirectional compression with provable improvements in total communication complexity as the number of workers increases. Theoretical findings align closely with empirical experiments, underscoring the efficiency of the proposed algorithms.<\/p>\n<p><strong>Bio: <\/strong>Kaja Gruntkowska\u00a0is a PhD student in Optimization for Machine Learning at KAUST, advised by Prof. Peter Richt\u00e1rik. Her research focuses on developing the algorithmic and mathematical foundations of randomized optimization, with a particular emphasis on distributed computing. She works on designing practically motivated algorithms with provable convergence guarantees, bridging theory and real-world applications to advance scalable machine learning. She completed her Bachelor&#8217;s in Mathematics and Statistics at the University of Warwick and earned a Master&#8217;s in Statistical Science from the University of Oxford.<\/p>\n<p class=\"p1\"><em>Please note that the meeting URL for this event can be seen on the cels-seminars website which requires an Argonne login.<\/em><\/p>\n<div class=\"tribe-events-single-event-description tribe-events-content\">\n<p>See all upcoming talks at\u00a0<a href=\"https:\/\/www.anl.gov\/mcs\/lans-seminars\">https:\/\/www.anl.gov\/mcs\/lans-seminars<\/a><\/p>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>Seminar Title: Improving the Worst-Case Bidirectional Communication Complexity for Nonconvex Distributed Optimization under Function Similarity Speaker: Kaja Gruntkowska, PhD Student, Optimization, Machine Learning, KAUST Date\/Time: Thursday, May 1, 2025\/ 11:00AM \u2013 12:00 PM (Virtual) Location:\u00a0See Meeting URL on the cels-seminars &hellip; <a href=\"https:\/\/wordpress.cels.anl.gov\/lans-seminars\/event\/3751\/\">Continue reading <span class=\"meta-nav\">&rarr;<\/span><\/a><\/p>\n","protected":false},"author":976,"featured_media":0,"template":"","meta":{"_acf_changed":false,"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"_tribe_events_status":"","_tribe_events_status_reason":"","footnotes":""},"tags":[],"tribe_events_cat":[2],"class_list":["post-3751","tribe_events","type-tribe_events","status-publish","hentry","tribe_events_cat-seminar","cat_seminar"],"acf":[],"_links":{"self":[{"href":"https:\/\/wordpress.cels.anl.gov\/lans-seminars\/wp-json\/wp\/v2\/tribe_events\/3751","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/wordpress.cels.anl.gov\/lans-seminars\/wp-json\/wp\/v2\/tribe_events"}],"about":[{"href":"https:\/\/wordpress.cels.anl.gov\/lans-seminars\/wp-json\/wp\/v2\/types\/tribe_events"}],"author":[{"embeddable":true,"href":"https:\/\/wordpress.cels.anl.gov\/lans-seminars\/wp-json\/wp\/v2\/users\/976"}],"version-history":[{"count":5,"href":"https:\/\/wordpress.cels.anl.gov\/lans-seminars\/wp-json\/wp\/v2\/tribe_events\/3751\/revisions"}],"predecessor-version":[{"id":3758,"href":"https:\/\/wordpress.cels.anl.gov\/lans-seminars\/wp-json\/wp\/v2\/tribe_events\/3751\/revisions\/3758"}],"wp:attachment":[{"href":"https:\/\/wordpress.cels.anl.gov\/lans-seminars\/wp-json\/wp\/v2\/media?parent=3751"}],"wp:term":[{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/wordpress.cels.anl.gov\/lans-seminars\/wp-json\/wp\/v2\/tags?post=3751"},{"taxonomy":"tribe_events_cat","embeddable":true,"href":"https:\/\/wordpress.cels.anl.gov\/lans-seminars\/wp-json\/wp\/v2\/tribe_events_cat?post=3751"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}