ADAPTIVE GRADIENT DESCENT WITHOUT DESCENT

von Malitsky, Yura
Veröffentlicht in International Conference on Machine Learning (37. : 2020 : Online) 37th International Conference on Machine Learning (ICML 2020) ; Part 9 of 15 (2021)

99% OF WORKER-MASTER COMMUNICATION IN DISTRIBUTED OPTIMIZATION IS NOT NEEDED

von MISHCHENKO, KONSTANTIN
Veröffentlicht in Conference on Uncertainty in Artificial Intelligence (36. : 2020 : Online) 36th Conference on Uncertainty in Artificial Intelligence (UAI 2020) ; Volume 3 of 3 (2021)

Proximal and Federated Random Reshuffling

von Mishchenko, Konstantin
Veröffentlicht in International Conference on Machine Learning (39. : 2022 : Baltimore, Md.; Online) International Conference on Machine Learning (ICML 2022) ; Part 19 of 33 (2023)

ProxSkip: Yes! Local Gradient Steps Provably Lead to Communication Acceleration! Finally!

von Mishchenko, Konstantin
Veröffentlicht in International Conference on Machine Learning (39. : 2022 : Baltimore, Md.; Online) International Conference on Machine Learning (ICML 2022) ; Part 19 of 33 (2023)

RANDOM RESHUFFLING: SIMPLE ANALYSIS WITH VAST IMPROVEMENTS

von Mishchenko, Konstantin
Veröffentlicht in NeurIPS (34. : 2020 : Online) 34th Conference on Neural Information Processing Systems (NeurIPS 2020) ; Volume 21 of 27 (2021)

Asynchronous SGD Beats Minibatch SGD Under Arbitrary Delays

von Mishchenko, Konstantin
Veröffentlicht in NeurIPS (36. : 2022 : New Orleans, La.; Online) 36th Conference on Neural Information Processing Systems (NeurIPS 2022 ; Volume 1 of 50 (2023)