1 paper · 1 filter
Aditya Balu, Sergio Botelho, Biswajit Khara +6
We consider the distributed training of large-scale neural networks that serve as PDE solvers producing full field outputs. We specifically consider neural solvers for the generali…