Eigenvalues of a tridiagonal block matrix in Pytorch

1

Hi,

what is the fastest way to calculate the eigenvalues of a tridiagonal block matrix in pytorch?

Ór

what is the fastest way to check whether a tridiagonal block matrix is postive semi-definite and calculate log det of it in pytorch?

is there an equivalent function (for tridiagonal block matrix) in Pytorch to scipy.linalg.eigvalsh_tridiagonal ?

Thanks!