ray.rllib.utils.exploration.stochastic_sampling.StochasticSampling.get_state

StochasticSampling.get_state(sess: tf.Session | None = None) → Dict[str, numpy.array | jnp.ndarray | tf.Tensor | torch.Tensor]

Returns the current exploration state.

Parameters:

sess – An optional tf Session object to use.

Returns:

The Exploration object’s current state.