Hi all. When using linfa-ensemble I noticed the two fitted-model types (EnsembleLearner<M> and AdaBoost<M, L>) cannot be currently persisted, even though the underlying model M is, e.g. EnsembleLearner<DecisionTree<F, L>>.
The extension of serialization to the ensemble types should involve, in principle, only derive bounds with no manual impl Serialize. I'm working on this in my fork and I would be happy to contribute if you want.
Hi all. When using
linfa-ensembleI noticed the two fitted-model types (EnsembleLearner<M>andAdaBoost<M, L>) cannot be currently persisted, even though the underlying modelMis, e.g.EnsembleLearner<DecisionTree<F, L>>.The extension of serialization to the ensemble types should involve, in principle, only derive bounds with no manual
impl Serialize. I'm working on this in my fork and I would be happy to contribute if you want.