baybe.surrogates.utils.batchify_ensemble_predictor

baybe.surrogates.utils.batchify_ensemble_predictor(base_predictor: Callable[[Tensor], Tensor])[source]

Wrap an ensemble predictor to make it evaluate t-batches as an augmented q-batch.

Parameters:

base_predictor (Callable[[Tensor], Tensor]) – The ensemble predictor to be wrapped.

Return type:

Callable[[Tensor], Tensor]

Returns:

The wrapped predictor.