Software Open Access
Stiller, Patrick;
Zhdanov, Maksim;
Rustamov, Jeyhun;
Bethke, Friedrich;
Hoffmann, Nico
Neural Solvers are neural network-based solvers for partial differential equations and inverse problems. The framework implements scalable physics-informed neural networks Physics-informed neural networks allow strong scaling by design. Therefore, we have developed a framework that uses data parallelism to accelerate the training of physics-informed neural networks significantly. To implement data parallelism, we use the Horovod framework, which provides near-ideal speedup on multi-GPU regimes.
Name | Size | |
---|---|---|
NeuralSolvers-v.0.1.zip
md5:e5c607ca66ca5486779e58f186e84d5b |
1.6 MB | Download |
All versions | This version | |
---|---|---|
Views | 684 | 684 |
Downloads | 18 | 18 |
Data volume | 28.1 MB | 28.1 MB |
Unique views | 530 | 530 |
Unique downloads | 18 | 18 |
Stiller, Patrick, Zhdanov, Maksim, Rustamov, Jeyhun, Bethke, Friedrich, & Hoffmann, Nico. (2021, September 6). Neural Solvers (Version 0.1). Rodare. http://doi.org/10.14278/rodare.1194