no code implementations • 2 May 2024 • Haiguang Li, Usama Pervaiz, Joseph Antognini, Michał Matuszak, Lawrence Au, Gilles Roux, Trausti Thormundsso
To address this, developers often face a trade-off between model accuracy and power consumption, employing either computationally intensive models on high-power cores or pared-down models on low-power cores.
no code implementations • 5 Apr 2024 • Haiguang Li, Usama Pervaiz, Michał Matuszak, Robert Kamara, Gilles Roux, Trausti Thormundsson, Joseph Antognini
The DSL builds upon the GC architecture, leveraging a dynamic pathway selection, and adapting model complexity in response to the innate structure of the data.
no code implementations • 8 Nov 2018 • Christopher J. Shallue, Jaehoon Lee, Joseph Antognini, Jascha Sohl-Dickstein, Roy Frostig, George E. Dahl
Along the way, we show that disagreements in the literature on how batch size affects model quality can largely be explained by differences in metaparameter tuning and compute budgets at different batch sizes.