Basically is there a Numpy or PyTorch function that does this:
vp_sa_s=mdp_data[\'sa_s\'].detach().clone() dims = vp_sa_s.size() for i in range(dims[0]): for