softmax_cross_entropy_with_logits#
- ray.rllib.utils.torch_utils.softmax_cross_entropy_with_logits(logits: NDArray[Any] | jnp.ndarray | tf.Tensor | torch.Tensor, labels: NDArray[Any] | jnp.ndarray | tf.Tensor | torch.Tensor) NDArray[Any] | jnp.ndarray | tf.Tensor | torch.Tensor[source]#
Same behavior as tf.nn.softmax_cross_entropy_with_logits.
- Parameters:
logits (NDArray[Any] | jnp.ndarray | tf.Tensor | torch.Tensor) – The input predictions.
labels (NDArray[Any] | jnp.ndarray | tf.Tensor | torch.Tensor) – The labels corresponding to
logits.
- Returns:
The resulting softmax cross-entropy given predictions and labels.
- Return type:
NDArray[Any] | jnp.ndarray | tf.Tensor | torch.Tensor