tf.keras.ops.sparse_categorical_crossentropy

Computes sparse categorical cross-entropy loss.

The sparse categorical cross-entropy loss is similar to categorical cross-entropy, but it is used when the target tensor contains integer class labels instead of one-hot encoded vectors. It measures the dissimilarity between the target and output probabilities or logits.

target The target tensor representing the true class labels as integers. Its shape should match the shape of the output tensor except for the last dimension.
output The output tensor representing the predicted probabilities or logits. Its shape should match the shape of the target tensor except for the last dimension.
from_logits (optional) Whether output is a tensor of logits or probabilities. Set it to True if output represents logits; otherwise, set it to False if output represents probabilities. Defaults toFalse.
axis (optional) The axis along which the sparse categorical cross-entropy is computed. Defaults to -1, which corresponds to the last dimension of the tensors.

Integer tensor: The computed sparse categorical cross-entropy loss between target and output.

Example:

target = keras.ops.convert_to_tensor([0, 1, 2], dtype=int32)
output = keras.ops.convert_to_tensor(
[[0.9, 0.05, 0.05],
 [0.1, 0.8, 0.1],
 [0.2, 0.3, 0.5]])
sparse_categorical_crossentropy(target, output)
array([0.10536056 0.22314355 0.6931472 ], shape=(3,), dtype=float32)