Zoe McCarthy
Zoe McCarthy
Ok I think I mostly understand. Basically you're doing a basis expansion of the Wiener process a la a Fourier expansion and then we solve the SDE in the coefficients...
Does it make sense to do something like stochastic sampling strategy during training but sample from the $N(0, 1)$'s instead of a uniform distribution? Matching the training sampling distribution to...
Generally I've found the best results with QuadratureTraining with ~1000-2000+ maxiters and StochasticTraining with 128-256+ samples. GridTraining should basically never be used. The number of observations for parameter estimation here...
This works! I want to do some cleanup of hardcoded log locations and make sure the interface is right before we merge and release it, so I'll request a review...
Yeah I'll clean it up and make it ready to merge after this next sprint.
I'm working on a project deadline for this Tuesday. I'll go through and add more detailed comments after that.
I agree with Chris that I think most of that work should not go here as it is mostly neural architecture related. I think it could make sense to add...
Also I think there's an issue with your line ending commit style, and that's why almost every line has a change. Are you committing in Windows line-ending style? I believe...
Also, here's an example of using the `additional_loss` interface for including your own loss terms in the PINN optimization: https://neuralpde.sciml.ai/stable/pinn/parm_estim/
I've been really busy with a project deadline on Tuesday, I should be able to do a thorough review and guide after that.