search

SciML / NeuralPDE.jl

Physics-Informed Neural Networks (PINN) Solvers of (Partial) Differential Equations for Scientific Machine Learning (SciML) accelerated simulation
https://docs.sciml.ai/NeuralPDE/stable/
Other
923 stars 196 forks source link

refactor: correctly lower quadrature training strategy in NNODE #836

Closed sathvikbhagavan closed 3 months ago

sathvikbhagavan commented 3 months ago

Along with correctly lowering quadrature training strategy in NNODE, things done:

  1. batch keyword is incorrectly described in NNODE docstring
  2. default tolerances were switched out for abstol and reltol in quadrature training.
  3. Lux.transform is deprecated for FromFluxAdaptor
ChrisRackauckas commented 3 months ago

It seems the tolerances just need to be changed?