Create a Python-based converter between PyTorch models and NetKet Machine

Created by: gcarleo

Something like

model = torch.nn.Sequential(
          torch.nn.Linear(D_in, H),
          torch.nn.ReLU(),
          torch.nn.Linear(H, D_out),
        ).to(device)

machine = nk.from_torch(model)

or maybe we can even define an automatic converter, such that model in this example can be passed directly to Vmc or other objects?