ray.rllib.models.distributions.Distribution.kl#

abstract Distribution.kl(other: Distribution, **kwargs) → numpy.array | jnp.ndarray | tf.Tensor | torch.Tensor[source]#

The KL-divergence between two distributions.

Parameters:

  • other – The other distribution.

  • **kwargs – Forward compatibility placeholder.

Returns:

The KL-divergence between the two distributions.