I have a function whose argument is the singular value of a matrix. Is it possible to get the derivative of this function using autograd? I am new to pyTorch and would appreciate some help regarding this issue.
yeah if you are using svd. it is differentiable.