Cannot convert "at::Tensor" to "nc::Ndarray" libtorch

Hi all, i am trying to deploy a project using libtorch.I have some post processing steps to do.The output of my model is a Tensor and i have to convert it to ndarray to continue with the post processing.In python this can be easily done with “tensor.numpy()” .Is there any equivalent for that in cpp?
I have tried various methods but none is working.
Please help me with this.

I think you can use<float>(), but what are you doing with this c++ numpy implementation that you can’t do directly in libtorch? Assuming you are using this, it doesn’t look like there is anything there that you couldn’t do while keeping your data as a tensor.