snn.losses

Base

class snn.losses.Loss

Bases: object

forward(y_pred, y_true)

backward(y_pred, y_true)

Regression

class snn.losses.MeanSquaredError

Bases: Loss

forward(y_pred, y_true)

backward(y_pred, y_true)

class snn.losses.MeanAbsoluteError

Bases: Loss

forward(y_pred, y_true)

backward(y_pred, y_true)

class snn.losses.HuberLoss(delta=1.0)

Bases: Loss

forward(y_pred, y_true)

backward(y_pred, y_true)

Classification

class snn.losses.BinaryCrossentropy(from_logits=False, eps=1e-07)

Bases: Loss

forward(y_pred, y_true)

backward(y_pred, y_true)

class snn.losses.CategoricalCrossentropy(from_logits=False, eps=1e-07)

Bases: Loss

forward(y_pred, y_true)

backward(y_pred, y_true)

class snn.losses.SparseCategoricalCrossentropy(from_logits=False, eps=1e-07)

Bases: Loss

forward(y_pred, y_true)

backward(y_pred, y_true)

Divergence

class snn.losses.KLDivergence(eps=1e-07)

Bases: Loss

forward(y_pred, y_true)

backward(y_pred, y_true)

Factory

snn.losses.get(identifier)

Return a Loss instance from a string, instance, or dict.

Parameters:

identifier (str, Loss, or dict) –

• String key — "mse", "categorical_crossentropy", …

• Loss instance — returned unchanged.

• Config dict — {"name": "huber", "delta": 2.0}. Any key accepted by the loss constructor may appear; "name" (or "class") selects the class.

Raises:

ValueError – Unknown string or dict name.