sniklaus/pytorch-unflow

loss function

tlok666 opened this issue · 1 comments

Hi, I read this code but find no loss function. Are you tried to train the network with this project?

I am afraid that I purely focused on replicating the original implementation in terms of inference. I have not made any attempt to replicate their work in terms of training. My implementation is fully differentiable though, which is why you just need to provide a data loader and implement a loss function to train it yourself.