baybe.surrogates.utils.batchify¶ baybe.surrogates.utils.batchify(posterior: Callable[[Surrogate, Tensor], tuple[Tensor, Tensor]])[source]¶ Wrap Surrogate posterior functions to enable proper batching. More precisely, this wraps model that are incompatible with t- and q-batching such that they become able to process batched inputs. Parameters: posterior (Callable[[Surrogate, Tensor], tuple[Tensor, Tensor]]) – The original posterior function. Return type: Callable[[Surrogate, Tensor], tuple[Tensor, Tensor]] Returns: The wrapped posterior function.