shim

Pytorch implementation of the tensor shim.

Classes

PyTorchBackendTensorShim

class PyTorchBackendTensorShim():

PyTorch backend shim/bridge for converting from/to PyTorch tensors.

Ancestors

• BackendTensorShim
• abc.ABC
• bitfount.types._BaseSerializableObjectMixIn

Variables

• static fields_dict : ClassVar[dict[str, marshmallow.fields.Field]]

• static nested_fields : ClassVar[dict[str, collections.abc.Mapping[str, Any]]]

Static methods

---

clamp_params

def clamp_params(    p: _TensorLike, prime_q: int, precision: int, num_workers: int,) ‑> bitfount.types._TensorLike:

Method for clipping params for secure sharing.

Constrains the parameters for secure sharing to be within the required range for secure sharing. Used only when steps_between_parameter_updates is 1.

Arguments

• p: The tensor to be constrained.
• prime_q: The prime use for secret aggregation.
• precision: The precision used for secret aggregation.
• num_workers: The number of workers taking part in the secure aggregation.

Returns The clamped parameters.

is_tensor

def is_tensor(p: Any) ‑> bool:

See base class.

to_list

def to_list(p: Union[np.ndarray, _TensorLike]) ‑> list:

See base class.

to_numpy

def to_numpy(t: Union[_TensorLike, list[float]]) ‑> numpy.ndarray:

See base class.

to_tensor

def to_tensor(p: Sequence, **kwargs: Any) ‑> bitfount.types._TensorLike:

See base class.

---