Registration is open for TensorFlow Dev Summit 2020 Learn more

tf.keras.losses.categorical_crossentropy

TensorFlow 1 version View source on GitHub

Computes the categorical crossentropy loss.

Aliases: tf.keras.metrics.categorical_crossentropy, tf.losses.categorical_crossentropy, tf.metrics.categorical_crossentropy

tf.keras.losses.categorical_crossentropy(
    y_true,
    y_pred,
    from_logits=False,
    label_smoothing=0
)

Args:

  • y_true: tensor of true targets.
  • y_pred: tensor of predicted targets.
  • from_logits: Whether y_pred is expected to be a logits tensor. By default, we assume that y_pred encodes a probability distribution.
  • label_smoothing: Float in [0, 1]. If > 0 then smooth the labels.

Returns:

Categorical crossentropy loss value.

Compat aliases