A question about directional loss.

[Bunnybeibei]

Thank you for your impressive work, especially some designs on losses. However, I found the sign function used in the directional loss to be non-derivable, resulting in zero gradients. How does this loss work in the final result?

[yhr91]

Thanks for your comments. Yes, you are correct that the sign function is not differentiable and this is an error. One way to circumvent this is to use the tanh function. We plan to update the repo once we've done some systematic tests but so far it looks like this improves performance slightly.

More discussion here: #36