BaseVQC#

class pyqit.models.layers.BaseVQC(n_qubits=4, n_layers=3, ansatz=<class 'pyqit.ansatzes.sel.SELAnsatz'>, encoder=<class 'pyqit.core.embeddings.AngleEmbedding'>, measure_fn=None, measure_wires=None, device='default.qubit', shots=None, diff_method='best')[source]#

Bases: BaseQuantumModel

An embedding, an ansatz, a measurement: the circuit the VQC models share.

Builds the ansatz and the embedding from their classes, draws the weights, and registers one QNode under main_circuit. It has no readout of its own, so you subclass it and never instantiate it directly. A subclass implements _resolve_readout(n_qubits, measure_fn, measure_wires), which sets _measure_fn and _measure_wires, and forward, which runs execute_qnode("main_circuit", X, **custom_weights) and maps the raw output. VQCClassifier, VQCRegressor and QuantumLayer differ only in those two methods.

Parameters:

  • n_qubits (int, default 4)

  • n_layers (int, default 3) – Depth passed to ansatz.

  • ansatz (type, default SELAnsatz) – Ansatz class, not an instance.

  • encoder (type, default AngleEmbedding) – Embedding class, not an instance. Stored as embedding_obj, which drives prescaling.

  • measure_fn (callable, optional) – Handed to _resolve_readout, which picks the default.

  • measure_wires (list of int, optional) – Handed to _resolve_readout, which picks the default.

  • device (str, default "default.qubit")

  • shots (int, optional)

  • diff_method (str, default "best") – Passed to the QNode. "best" picks backprop on a simulator; "parameter-shift" rehearses a hardware run’s gradient cost.

diff_methods(X) → dict#

Differentiation method per QNode.

Parameters:

X (array-like) – One prescaled batch; only its shape matters.

Returns:

QNode name to method name.

Return type:

dict

execute_qnode(name: str, X, **custom_weights)#

Run the QNode or dense layer registered under name on a batch.

Parameters:

  • name (str) – Name passed to register_qnode.

  • X (array-like)

  • **custom_weights – Flat “<name>.<weight>” overrides; unprefixed keys are ignored. Falls back to the model’s own weights when empty.

Return type:

array-like

abstractmethod forward(X)#

Run the model on a batch and return its raw output.

get_interface()#

PennyLane QNode interface for the active backend.

init_weights(weight_shapes: dict) → dict#

Draw uniform [0, 1) starting weights from numpy’s global RNG.

The range matches Qiskit ML’s default initial_point and PennyLane’s template examples; uniform [0, 2pi) is the Haar-like regime where gradients vanish (McClean et al. 2018).

Call this before building the device: a PennyLane device seeded "global" consumes numpy’s RNG at construction by a device-dependent amount, so weights drawn after it differ per device for the same seed.

Parameters:

weight_shapes (dict) – Weight name to shape, as returned by an ansatz’s get_weight_shapes.

Return type:

dict

is_fitted() → bool#

Whether Trainer.fit has trained this model.

register_dense(name: str, n_in: int, n_out: int, weights=None)#

Register a classical dense layer X @ weight.T + bias under name.

Lives in the same registry as the QNodes, so weights, update_weights, checkpoints and the flat-kwargs routing cover it with no further plumbing. Run it with execute_qnode.

Parameters:

  • name (str)

  • n_in (int)

  • n_out (int)

  • weights (dict, optional) – {"weight", "bias"} from init_dense_weights; drawn when omitted.

register_qnode(name: str, qnode: QNode, weight_shapes: dict, weights=None)#

Wrap qnode for the active backend and store it under name.

Parameters:

  • name (str) – Key under which the node’s weights appear in weights.

  • qnode (qml.QNode)

  • weight_shapes (dict) – Weight name to shape, as returned by an ansatz’s get_weight_shapes.

  • weights (dict, optional) – Starting weights from init_weights; drawn here when omitted, so both backends start from the same point for the same seed.

update_weights(flat_weights_dict)#

Write flat_weights_dict into the model’s own weights.

No-op under torch, where autograd owns the nn.Parameter objects directly.

Parameters:

flat_weights_dict (dict) – Keyed like weights.

weight_groups() → dict#

weights keys by group, "quantum" (QNodes) and "classical".

Empty groups are omitted. The training loops build one optimizer per group, which is what lets Trainer(learning_rate={...}) set a rate per group.

property weights#

Flat {"<qnode_name>.<weight_name>": array} dict, both backends.