Lstsq for rank deficient matrix?

Hello,

Is there a version of torch.lstsq function that works for rank deficient matrices, similar to numpy’s linalg.lstsq?