jasmine.metrics.categorical_cross_entropy

jasmine.metrics.categorical_cross_entropy(y_true, y_pred, from_logits: bool = False)[source]

Compute the Categorical Cross-Entropy loss between true and predicted values.

Parameters:

y_true (jnp.ndarray) – True categorical target values (one-hot encoded).
y_pred (jnp.ndarray) – Predicted probabilities or logits for each class.
from_logits (bool) – If True, y_pred is expected to be a raw logit output.

Returns:

Computed Categorical Cross-Entropy loss.

Return type: