Synergize with qiboml developments

[chmwzc]

VQE-related methods will be moving from the main qibo repository to qiboml. To integrate qibochem code with this development

[chmwzc]

Update: Tried following the VQE tutorial on qiboml, with some modifications

Code block

 import numpy as np
import torch.optim as optim
from qibo import set_backend
from qiboml.models.decoding import Expectation
from qiboml.interfaces.pytorch import QuantumModel
from qibochem.ansatz import ucc_ansatz
from qibochem.driver import Molecule

def main():
    # Chemistry stuff
    mol = Molecule(xyz_file="../h2.xyz")
    mol.run_pyscf()
    hamiltonian = mol.hamiltonian()
    circuit = ucc_ansatz(mol)
    circuit.draw()
    # qiboml stuff
    set_backend("qiboml", platform="pytorch")
    decoding = Expectation(nqubits=hamiltonian.nqubits, observable=hamiltonian)
    model = QuantumModel(
        circuit_structure=circuit,
        decoding=decoding,
    )
    print("Exact ground state: ", min(hamiltonian.eigenvalues()))
    optimizer = optim.Adam(model.parameters(), lr=0.05)
    for iteration in range(300):
        optimizer.zero_grad()
        cost = model()
        cost.backward()
        optimizer.step()
        if iteration % 20 == 0:
            print(f"Iteration {iteration}: Cost = {cost.item():.6f}")

if __name__ == "__main__":
    main()

Ran into an error (see below). Will investigate/continue monitoring

Error

[Qibo 0.2.18|ERROR|2025-06-20 08:54:05]: Cannot calculate Hamiltonian expectation value for state of type <class 'torch.Tensor'>
Traceback (most recent call last):
  File "/home/wongzc/testing/qiboml/test.py", line 53, in <module>
    main()
  File "/home/wongzc/testing/qiboml/test.py", line 44, in main
    cost = model()
           ^^^^^^^
  File "/home/wongzc/miniforge3/envs/ml-chem/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1751, in _wrapped_call_impl
    return self._call_impl(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/wongzc/miniforge3/envs/ml-chem/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1762, in _call_impl
    return forward_call(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/wongzc/miniforge3/envs/ml-chem/lib/python3.12/site-packages/qiboml/interfaces/pytorch.py", line 81, in forward
    x = self.decoding(circuit)
        ^^^^^^^^^^^^^^^^^^^^^^
  File "/home/wongzc/miniforge3/envs/ml-chem/lib/python3.12/site-packages/qiboml/models/decoding.py", line 159, in __call__
    return self.observable.expectation(
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/wongzc/miniforge3/envs/ml-chem/lib/python3.12/site-packages/qibo/hamiltonians/hamiltonians.py", line 484, in expectation
    return Hamiltonian.expectation(self, state, normalize)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/wongzc/miniforge3/envs/ml-chem/lib/python3.12/site-packages/qibo/hamiltonians/hamiltonians.py", line 117, in expectation
    raise_error(
  File "/home/wongzc/miniforge3/envs/ml-chem/lib/python3.12/site-packages/qibo/config.py", line 46, in raise_error
    raise exception(message)
TypeError: Cannot calculate Hamiltonian expectation value for state of type <class 'torch.Tensor'>

[chmwzc]

Remember to add qiboml as a dependency when we start using it here as well