torchdms.loss.diff_penalty

torchdms.loss.diff_penalty(betas)[source]

Computes l1 norm of the difference between adjacent betas for each latent dimension.