Closed mw66 closed 1 week ago
Hi the training loss is by default MSE, but the saved results are for RMSE. RMSE loss is not differentiable (at zero) hence would have optimization problems. MSE has better landscape.
Hi the training loss is by default MSE, but the saved results are for RMSE. RMSE loss is not differentiable (at zero) hence would have optimization problems. MSE has better landscape.
OK, but can you doc this info, otherwise it's confusing when comparing the above code with the paper / tutorials.
https://github.com/KindXiaoming/pykan/blob/915726a6770bf439021f1bac38bfe39fd424159e/kan/KAN.py#L844
Also can you add the doc for this
loss_fn
param?