tfm.uplift.types.TwoTowerTrainingOutputs

Training tensors computed by a TwoTowerUpliftNetwork layer.

Inherits From: TwoTowerPredictionOutputs, TwoTowerNetworkOutputs

tfm.uplift.types.TwoTowerTrainingOutputs(
    shared_embedding,
    control_logits,
    treatment_logits,
    control_predictions,
    treatment_predictions,
    uplift,
    true_logits,
    true_predictions,
    is_treatment
)

true_logits logits for either the control or treatment group, depending on the corresponding value in the is_treatment tensor. It will contain treatment group logits for the is_treatment == 1 entries and control group logits otherwise.
true_predictions predictions for either the control or treatment group, depending on the corresponding value in the is_treatment tensor. It will contain treatment group predictions for the is_treatment == 1 entries and control group predictions otherwise.
is_treatment a boolean tf.Tensor indicating if the example belongs to the treatment group (True) or control group (False).
dtype

shape

Child Classes

class Spec

Methods

__eq__

__eq__(
    other
)

Return self==value.

__ne__

__ne__(
    other
)

Return self!=value.