ensemble_loss#
- hybrid_aif.train.ensemble_loss(members, batch, *, member_weights=None)[source]#
(total, EnsembleLosses): the summed member NLL, plus the per-member vector.Pure and differentiable in
members; each member’s NLL isdiag_nll()(the diagonal form, which keeps gradients finite under jit).member_weightsis an(M,)weight per member (default ones; a zero removes that member’s gradient); any other shape raisesValueError. The reduction is a sum, which matters where adam is not scale-free (epsand the clip).