API REFERENCE/omnibias-torch

omnibias.torch.blocks.linear

0.5.0rc1Release candidatePython >=3.10Apache-2.0

PyTorch activation derivatives, trainable operator layers, network jets, and direct differential-operator contractions.

import omnibias.torch.blocks.linear

Module contract & mathematical conventions

Operator-typed linear layer: cmbLinear.

A drop-in for torch.nn.Linear followed by an OperatorBlock on the output channels.

Shared formulas do not imply universal floating-point identity. Verify dtype, shape, derivative-order support, and the stated numerical assumptions for the path you use.

Module source ↗Source revision 9a2d73df0573Running reference examples →

CLASS

cmbLinear

omnibias.torch.cmbLinearSource ↗

class cmbLinear(nn.Module)

nn.Linear + per-channel OperatorBlock.

Parameters

in_features : int out_features : int op : OpName, default "identity" Operator role for the output activation. base : str or ActivationSpec, default "sigmoid" bias : bool, default True Linear-layer bias (independent of the OMBU bias terms). block_kwargs : dict, optional Extra keyword arguments passed to OperatorBlock.

Parameters

Name	Type & default	Contract
in_featurespositional-or-keyword	intRequired	int
out_featurespositional-or-keyword	intRequired	int
oppositional-or-keyword	OpNameDefault: 'identity'	OpName, default "identity" Operator role for the output activation.
basepositional-or-keyword	str | ActivationSpec[Tensor]Default: 'sigmoid'	str or ActivationSpec, default "sigmoid"
biaspositional-or-keyword	boolDefault: True	bool, default True Linear-layer bias (independent of the OMBU bias terms).
block_kwargspositional-or-keyword	dict[str, Any] | NoneDefault: None	dict, optional Extra keyword arguments passed to OperatorBlock.

Executed regression · cmblinear forward shape

Recorded coverage confirms this test executed the API body. This example belongs to the pinned upstream test suite. It may use fixtures and imports from its full file. Run it from that checkout with its test dependencies installed.

Full file, imports & fixtures ↗

def test_cmblinear_forward_shape() -> None:
    layer = cmbLinear(5, 3, op="identity", base="tanh")
    out = layer(torch.randn(2, 5))
    assert out.shape == (2, 3)

python -m pytest "packages/omnibias-torch/tests/test_blocks.py::test_cmblinear_forward_shape" -q

Regression example · cmblinear uses default dtype

This example belongs to the pinned upstream test suite. It may use fixtures and imports from its full file. Run it from that checkout with its test dependencies installed.

Full file, imports & fixtures ↗

def test_cmblinear_uses_default_dtype(torch_default_float64) -> None:
    fc = cmbLinear(in_features=8, out_features=4, op="grad", base="tanh")
    assert fc.linear.weight.dtype == torch.float64
    assert fc.block.ombu.biases.dtype == torch.float64

python -m pytest packages/omnibias-torch/tests/test_audit_regressions.py -k test_cmblinear_uses_default_dtype -q

Fields & methods

__init__method

__init__(self, in_features: int, out_features: int, op: OpName='identity', base: str | ActivationSpec[Tensor]='sigmoid', bias: bool=True, block_kwargs: dict[str, Any] | None=None) -> None

See the class contract above and the exact method implementation below.

Parameters

Name	Type & default	Contract
in_featurespositional-or-keyword	intRequired	See the operation contract and implementation below.
out_featurespositional-or-keyword	intRequired	See the operation contract and implementation below.
oppositional-or-keyword	OpNameDefault: 'identity'	See the operation contract and implementation below.
basepositional-or-keyword	str | ActivationSpec[Tensor]Default: 'sigmoid'	See the operation contract and implementation below.
biaspositional-or-keyword	boolDefault: True	See the operation contract and implementation below.
block_kwargspositional-or-keyword	dict[str, Any] | NoneDefault: None	See the operation contract and implementation below.

Returns: None

Executed regression · cmblinear forward shape

Recorded coverage confirms this test executed the API body. This example belongs to the pinned upstream test suite. It may use fixtures and imports from its full file. Run it from that checkout with its test dependencies installed.

Full file, imports & fixtures ↗

def test_cmblinear_forward_shape() -> None:
    layer = cmbLinear(5, 3, op="identity", base="tanh")
    out = layer(torch.randn(2, 5))
    assert out.shape == (2, 3)

python -m pytest "packages/omnibias-torch/tests/test_blocks.py::test_cmblinear_forward_shape" -q

Method implementation

    def __init__(
        self,
        in_features: int,
        out_features: int,
        op: OpName = "identity",
        base: str | ActivationSpec[Tensor] = "sigmoid",
        bias: bool = True,
        block_kwargs: dict[str, Any] | None = None,
    ) -> None:
        super().__init__()
        self.in_features = in_features
        self.out_features = out_features
        self.linear = nn.Linear(in_features, out_features, bias=bias)
        self.block = OperatorBlock(
            op=op,
            base=base,
            channels=out_features,
            **(block_kwargs or {}),
        )

View source ↗

forwardmethod

forward(self, x: Tensor) -> Tensor

See the class contract above and the exact method implementation below.

Parameters

Name	Type & default	Contract
xpositional-or-keyword	TensorRequired	See the operation contract and implementation below.

Returns: Tensor

Executed regression · cmblinear forward shape

Recorded coverage confirms this test executed the API body. This example belongs to the pinned upstream test suite. It may use fixtures and imports from its full file. Run it from that checkout with its test dependencies installed.

Full file, imports & fixtures ↗

def test_cmblinear_forward_shape() -> None:
    layer = cmbLinear(5, 3, op="identity", base="tanh")
    out = layer(torch.randn(2, 5))
    assert out.shape == (2, 3)

python -m pytest "packages/omnibias-torch/tests/test_blocks.py::test_cmblinear_forward_shape" -q

Method implementation

    def forward(self, x: Tensor) -> Tensor:
        out: Tensor = self.block(self.linear(x))
        return out

View source ↗

extra_reprmethod

extra_repr(self) -> str

See the class contract above and the exact method implementation below.

Returns: str

Method implementation

    def extra_repr(self) -> str:
        return (
            f"in_features={self.in_features}, out_features={self.out_features}, "
            f"op={self.block.op!r}, base={self.block.base_name!r}"
        )

View source ↗

in_featuresattribute

in_features: int = in_features

int

View source ↗

out_featuresattribute

out_features: int = out_features

int

View source ↗

linearattribute

linear = nn.Linear(in_features, out_features, bias=bias)

Public instance attribute assigned by __init__. The source assignment is linear = nn.Linear(in_features, out_features, bias=bias); it describes instance state, not a literal constructor default.

Regression example · cmblinear uses default dtype

This example belongs to the pinned upstream test suite. It may use fixtures and imports from its full file. Run it from that checkout with its test dependencies installed.

Full file, imports & fixtures ↗

def test_cmblinear_uses_default_dtype(torch_default_float64) -> None:
    fc = cmbLinear(in_features=8, out_features=4, op="grad", base="tanh")
    assert fc.linear.weight.dtype == torch.float64
    assert fc.block.ombu.biases.dtype == torch.float64

python -m pytest packages/omnibias-torch/tests/test_audit_regressions.py -k test_cmblinear_uses_default_dtype -q

View source ↗

blockattribute

block = OperatorBlock(op=op, base=base, channels=out_features, **block_kwargs or {})

Public instance attribute assigned by __init__. The source assignment is block = OperatorBlock(op=op, base=base, channels=out_features, **block_kwargs or {}); it describes instance state, not a literal constructor default.

Regression example · cmblinear uses default dtype

This example belongs to the pinned upstream test suite. It may use fixtures and imports from its full file. Run it from that checkout with its test dependencies installed.

Full file, imports & fixtures ↗

def test_cmblinear_uses_default_dtype(torch_default_float64) -> None:
    fc = cmbLinear(in_features=8, out_features=4, op="grad", base="tanh")
    assert fc.linear.weight.dtype == torch.float64
    assert fc.block.ombu.biases.dtype == torch.float64

python -m pytest packages/omnibias-torch/tests/test_audit_regressions.py -k test_cmblinear_uses_default_dtype -q

View source ↗

Read the implementation

class cmbLinear(nn.Module):
    """``nn.Linear`` + per-channel :class:`OperatorBlock`.

    Parameters
    ----------
    in_features : int
    out_features : int
    op : :class:`OpName`, default ``"identity"``
        Operator role for the output activation.
    base : str or :class:`ActivationSpec`, default ``"sigmoid"``
    bias : bool, default True
        Linear-layer bias (independent of the OMBU bias terms).
    block_kwargs : dict, optional
        Extra keyword arguments passed to :class:`OperatorBlock`.
    """

    def __init__(
        self,
        in_features: int,
        out_features: int,
        op: OpName = "identity",
        base: str | ActivationSpec[Tensor] = "sigmoid",
        bias: bool = True,
        block_kwargs: dict[str, Any] | None = None,
    ) -> None:
        super().__init__()
        self.in_features = in_features
        self.out_features = out_features
        self.linear = nn.Linear(in_features, out_features, bias=bias)
        self.block = OperatorBlock(
            op=op,
            base=base,
            channels=out_features,
            **(block_kwargs or {}),
        )

    def forward(self, x: Tensor) -> Tensor:
        out: Tensor = self.block(self.linear(x))
        return out

    def extra_repr(self) -> str:
        return (
            f"in_features={self.in_features}, out_features={self.out_features}, "
            f"op={self.block.op!r}, base={self.block.base_name!r}"
        )

Related APIsomnibias.torch.OperatorBlock↗

Other public import paths

• omnibias.torch.cmbLinear
• omnibias.torch.blocks.cmbLinear
• omnibias.torch.blocks.linear.cmbLinear

← Package overviewRunnable learning examples →