Losses

Losses are concrete callable types, following the Lux loss API: they take (model, ps, st, data) and return (scalar_loss, updated_state, NamedTuple()).

Prediction Shape

All losses expect 3D predictions: (outsize, K, batch) where K is the ensemble size. 2D outputs are reshaped to (outsize, 1, batch) automatically.

reduce_pred averages over K on raw predictions before any transformation.

Usage

MSE()                    # callable loss functor
LossType(:mse)           # same, from a symbol (`MSE()`)
MSE()(model, ps, st, data)

NeuroTabRegressor / NeuroTabClassifier still take loss=:mse (a symbol). fit converts that once to a functor and stores it on NeuroTabModel.loss.

Supported Losses

Symbol	Type	Pred shape	Target	Notes
:mse	MSE	(1, K, B)	scalar
:mae	MAE	(1, K, B)	scalar
:logloss	LogLoss	(1, K, B)	{0, 1}	raw logits
:mlogloss	MLogLoss	(C, K, B)	{1, …, C}	raw logits
:gaussian_mle	GaussianMLE	(2, K, B)	scalar	pred[1,:,:] = μ, pred[2,:,:] = log-σ
:tweedie	Tweedie	(1, K, B)	non-negative	log-scale pred, ρ = 1.5
:pearson	Pearson	(1, K, B)	scalar	negative Pearson correlation

Data Tuples

Tuple	Contents
(x, y)	standard training
(x, y, w)	weighted training
(x, y, w, offset)	with offset (e.g. boosting)

Signature

loss(model, ps, st, data) → (scalar_loss, updated_state, NamedTuple())