Svds in pytorch

Hi, in Matlab there is a function called “svds”, which finds the k largest eigenvalues of a matrix.
Is there any parallel function in pytorch (not svd)?

u, s, v = torch.svd(a)

s[:k] contains the k largest singular values of a. I don’t think svds exists in Pytorch.

1 Like

Thanks for the response!

My goal is to find only the k largest values in order to short my running time.
So, is there any other way to find them?