kosmos.ml.models.vqc.encoding

Classes

class VQCEncoding(input_dim: int, output_dim: int)

Bases: abc.ABC

Feature encoding for the VQC.

Initialize the encoding.

Parameters:

• input_dim (int) – The input dimension of the model.

• output_dim (int) – The output dimension of the model.

Properties

property num_qubits

Number of required qubits.

Methods

apply_operation(features: torch.Tensor, wires: collections.abc.Sequence[int]) → None

Apply the PennyLane operation for encoding.

Parameters:

• features (Tensor) – Input features.

• wires (Sequence[int]) – Target wires.

---

class AngleEmbedding(input_dim: int, output_dim: int, rotation: Literal['X', 'Y', 'Z'])

Bases: VQCEncoding

Angle embedding for the VQC.

Initialize the angle embedding.

Parameters:

• input_dim (int) – The input dimension of the model.

• output_dim (int) – The output dimension of the model.

• rotation (Literal["X", "Y", "Z"]) – The rotation to use for the angle embedding.

Properties

property num_qubits

Number of required qubits.

Methods

apply_operation(features: torch.Tensor, wires: collections.abc.Sequence[int]) → None

Apply the angle embedding operation.

Parameters:

• features (Tensor) – Input features.

• wires (Sequence[int]) – Target wires.

---

class AmplitudeEmbedding(input_dim: int, output_dim: int, pad_with: complex, *, normalize: bool)

Bases: VQCEncoding

Amplitude embedding for the VQC.

Initialize the amplitude embedding.

Parameters:

• input_dim (int) – The input dimension of the model.

• output_dim (int) – The output dimension of the model.

• pad_with (complex) – The input is padded with this constant to size \(2^n\).

• normalize (bool) – Whether to normalize the features.

Properties

property num_qubits

Number of required qubits.

Methods

apply_operation(features: torch.Tensor, wires: collections.abc.Sequence[int]) → None

Apply the amplitude embedding operation.

Parameters:

• features (Tensor) – Input features.

• wires (Sequence[int]) – Target wires.