1 paper · 1 filter
Linxuan Pan, Shucheng Yang, Junchi Yang
Distributed learning with multiple local steps reduces communication but aggravates drift, and its practical performance often depends on carefully tuned hyperparameters under hete…