Helper method that performs torch.cat on the shares of the Tensors. Args: session_uuid_str (str): UUID to identify the session on each pa

Operation to compute local z share and add mask to it. Args: x (ReplicatedSharedTensor): Secret. y (ReplicatedSharedT

Add the primitives required for the "conv2d" operation to the CryptoStore. Args: store (Any): The CryptoStore. primitives (Iterab

Retrieve the primitives from the CryptoStore. Those are needed for executing the "conv2d" operation. Args: store: the CryptoStore

Add the primitives required for the "conv_transpose2d" operation to the CryptoStore. Args: store (Any): The CryptoStore. primitiv

Retrieve the primitives from the CryptoStore. Those are needed for executing the "conv_transpose2d" operation. Args: store: the Cryp

Count wraps random. The Trusted Third Party (TTP) or Crypto provider should generate: - a set of shares for a random number - a set of s

Decompose a tenstor into its binary representation. Args: tensor (torch.Tensor): Tensor to decompose. ring_size (int): Ring size.

Distribute a list of shares. Args: shares (List[ShareTensor): list of shares to distribute. session (Session): Sessio

Apply the "div" operation between "self" and "y". Args: y (Union[int, float, torch.Tensor, "ShareTensor"]): Denominator.

From CrypTen Privately computes the number of wraparounds for a set a shares. To do so, we note that: [theta_x] = theta_z + [beta_xr] - [

Equal operator. Check if "self" is equal with another object given a set of attributes to compare. Args: y (Any): Object

Equal operator. Args: x1 (MPCTensor): First private value. x2 (MPCTensor): Second private value. Returns:

Evaluate the FSS protocol on the masked and public input `x_masked`. Args: session (Session): MPC Session b: rank of the evaluato

Falcon mask. Args: x_sh (ReplicatedSharedTensor): X share y_sh (ReplicatedSharedTensor) : Y share op_str

Perform the feedforward and compute the result for the implemented operation. Args: ctx (Dict[str, Any]): Context used to save in

Perform the feedforward and compute the result for the addition operation. Args: ctx (Dict[str, Any]): Context used to save infor

Perform the feedforward and compute the result for the summation operation. Args: ctx (Dict[str, Any]): Context used to save info

Perform the feedforward and compute the result for the sigmoid operation. Args: ctx (Dict[str, Any]): Context used to save inform

Perform the feedforward and compute the result for the substraction operation. Args: ctx (Dict[str, Any]): Context used to save i

Perform the feedforward and compute the result for the multiplication operation. Args: ctx (Dict[str, Any]): Context used to save

Perform the feedforward and compute the result for the multiplication operation. Args: ctx (Dict[str, Any]): Context used to save

Perform the feedforward and compute the result for the division operation. Args: ctx (Dict[str, Any]): Context used to save infor

Perform the feedforward and compute the result for the matrix multiplication operation. Args: ctx (Dict[str, Any]): Context used

Perform the feedforward and compute the result for the multiplication operation. Args: ctx (Dict[str, Any]): Context used to save

Perform the feedforward and compute the result for the conv2d operation. Args: ctx (Dict[str, Any]): Context used to save informa

Perform the feedforward and compute the result for the reshape operation. Args: ctx (Dict[str, Any]): Context used to save inform

Perform the feedforward and compute the result for the ReLU operation. Args: ctx (Dict[str, Any]): Context used to save informati

Perform the feedforward and compute the result for the MaxPool2D operation. Args: ctx (Dict[str, Any]): Context used to save info

Do a feedforward through the layer. Args: x (MPCTensor): the input Returns: An MPCTensor that results by app

Do a feedforward through the layer. Args: x (MPCTensor): the input Returns: An MPCTensor representing the la

Do a feedforward through the layer. Args: x (MPCTensor): the input Returns: An MPCTensor representing the la

Greater or equal operator. Args: other (MPCTensor): MPCTensor to compare. Returns: MPCTensor: Result of the

Get shares. Returns: List[torch.Tensor]: List of shares.

Get a torch module from a given MPC Layer module. The parameters of the models are not set. Args: linear_module (Linear)

Get the beaver triples for the conv2d operation. Args: *args: Arguments for _get_triples. **kwargs: Keyword arguments for _get_tr

Get the beaver triples for the matmul operation. Args: *args (List[ShareTensor]): Named arguments of :func:`beaver.__get_triples`.

Get the beaver triples for the multiplication operation. Args: *args (List[ShareTensor]): Named arguments of :func:`beaver.__get_triples`

Get the beaver triples for the conv_transpose2d operation. Args: *args: Arguments for _get_triples. **kwargs: Keyword arguments f

Greater than operator. Args: y: self>y Raises: NotImplementedError: Raised when implementation not present

Helper function that would compute the difference between all the elements in a tensor. Credits goes to the CrypTen team. Args: sess

Function that runs at each party for preparing the share. Reshape each share tensor to prepare it for calling 'argmax'. The new share would h

Hook a framework method such that we know how to treat it given that we call it. Ex: * if we call "numel" we want to call it on the

Hook a framework method such that we know how to treat it given that we call it. Ex: * if we call "numel" we want to call it on the

Hook a framework method. Ex: * if we call "numel" we want to forward it only to one share and return the result (without wra

Hook a framework property (only getter). Ex: * if we call "shape" we want to call it on the underlying tensor and return the

Hook a framework property (only getter). Ex: * if we call "shape" we want to call it on the underlying tensor and return the

Hook a framework property (only getter). Ex: * if we call "shape" we want to call it on the underlying share and return the

Set the same event loop to other threads/processes. This is needed because there are new threads/processes started with the Executor

Lower equal operator. Args: x1 (MPCTensor): First private value. x2 (MPCTensor): Second private value. Retur

Lower than operator. Args: y: self<y Raises: NotImplementedError: Raised when implementation not present

Lower than operator. Args: other (MPCTensor): MPCTensor to compare. Returns: MPCTensor: Result of the compar

Mask the private inputs. Add the share of alpha (the mask) that is held in the crypto store to the difference x1 - x2. As we aim at compa

Apply the "matmul" operation between "self" and "y". Args: y (Union[int, float, torch.Tensor, "ReplicatedSharedTensor"]): self@y

Add the primitives required for the "matmul" operation to the CryptoStore. Args: store (Any): The CryptoStore. primitives (Iterab

Retrieve the primitives from the CryptoStore. Those are needed for executing the "matmul" operation. Args: store (Dict[Tuple[int, in

Compute the maximum value for an MPCTensor. Args: x (MPCTensor): MPCTensor to be computed the maximum value on. dim (Optional[Uni

Compute the max pool for a tensor with 2 dimension. Args: x (MPCTensor): the MPCTensor on which to apply the operation kernel_siz

Helper function for the backwards step for max_pool2d. Credits goes to the CrypTen team. Args: grad (MPCTensor): gradient that comes

Helper function to compute the gradient needed to be passed to the parent node. Args: input_shape (Tuple[int]): the size of the input ten

Apply the "mul" operation between "self" and "y". Args: y (Union[int, float, torch.Tensor, "ReplicatedSharedTensor"]): self*y

Apply the "mul" operation between "self" and "y". Args: y (Union[int, float, torch.Tensor, "ShareTensor"]): self * y Ret

Function that is executed by the orchestrator to multiply two secret values. Args: x (MPCTensor): First value to multiply with. y

SPDZ Multiplication. Args: session_uuid_str (str): UUID to identify the session on each party side. eps (torch:tensor): Epsilon v

Add the primitives required for the "mul" operation to the CryptoStore. Arguments: store (Dict[Any, Any]): the CryptoStore primit

Retrieve the primitives from the CryptoStore. Those are needed for executing the "mul" operation. Args: store (Dict[Tuple[int, int],

Computes multiplication(x*y) modulo PRIME_NUMBER constant. Args: x (torch.Tensor): input tensor y (torch.tensor): inp

Not Equal operator. Args: y: self!=y Raises: NotImplementedError: Raised when implementation not present

Not equal operator. Args: other (MPCTensor): MPCTensor to compare. Returns: MPCTensor: Result of the compari

Get the parameters of the Linear module. Args: recurse (bool): For the moment not used. TODO Yields: Each pa

Get the parameters of the Linear module. Args: recurse (bool): For the moment not used. TODO Yields: Each pa

Get the precision for the FixedPrecision Encoder. Returns: int: precision.

This function will generate the values: [((0, 0, 0, 0), (0, 0, 0, 0), ...), ((1, 1, 1, 1), (1, 1, 1, 1)), ...]

Function that is executed by the orchestrator to divide a secret by a public value. Args: x (MPCTensor): Private numerator. y (Un

Get back the shares from all the parties and reconstruct the underlying value. Returns: Neural network module with the secret reconstruct

Reconstruct value from shares. Args: share_ptrs (List[ReplicatedSharedTensor]): List of RSTensor pointers. security_t

Reconstruct original value from shares. Args: share_ptrs (List[ShareTensor]): List of sharetensors. get_shares (boole

Reconstruct the shared state dict. Returns: The reconstructed state dict (Dict[str, Any])