boax.core.optimizers.initializers.q_batch_nonnegative

boax.core.optimizers.initializers.q_batch_nonnegative#

boax.core.optimizers.initializers.q_batch_nonnegative(samples, num_restarts, eta=1.0, alpha=0.0001)#

Q batch initializer.

Example

>>> initializer = q_batch_nonnegative(fun, samples, num_restarts)
>>> candidates = initializer(key)
Parameters:

  • fun – The scoring function.

  • samples (Union[Array, ndarray, bool, number]) – The candidate samples.

  • num_restarts (int) – The number of restarts.

  • eta (Union[Array, ndarray, bool, number, float, int]) – The temperature parameter.

  • alpha (Union[Array, ndarray, bool, number, float, int]) – The alpha parameter.

Return type:

Initializer

Returns:

The q-batch non-negative Initializer.