Classical Utilities¶

This package contains classical utilities that are used by both classical and quantum models.

str_to_activation¶

classical_utils.str_to_activation() → torch.nn.modules.module.Module¶

Converts string to PyTorch’s activation. Currently:

Output classes corresponding to input strings¶
activation	output
`relu`	`nn.ReLU()`
`sigmoid`	`nn.Sigmoid()`
`tanh`	`nn.Tanh()`
`leaky relu`, `leaky-relu`, `leaky_relu`	`nn.LeakyReLU()`
`angle`, `to-angle`, `range`, `to-range`	`ToRange`

Parameters: activation (str) – String representation of activation
Raises: ValueError – Error is thrown if activation string is not recognized.
Returns: PyTorch module for corresponding activation function with default values.
Return type: nn.Module

ToRange¶

class squid.models.classical_utils.ToRange(input_range: Tuple[Union[torch.Tensor, int, float], Union[torch.Tensor, int, float]] = (0, tensor([3.1416])))[source]¶

Bases: torch.nn.modules.module.Module

ToRange docs

Parameters: input_range (Tuple[ Union[torch.Tensor, int, float], Union[torch.Tensor, int, float] ], optional) – [description], defaults to (0, torch.Tensor([pi]))