NLLLoss

class ampligraph.latent_features.NLLLoss(loss_params={}, verbose=False)

Negative Log-Likelihood loss.

As described in [TWR+16].

\[\mathcal{L}(\Theta) = \sum_{t \in \mathcal{G} \cup \mathcal{C}}log(1 + exp(-y \, f_{model}(t;\Theta)))\]

where \(y \in \{-1, 1\}\) is the label of the statement, \(\mathcal{G}\) is the set of positives, \(\mathcal{C}\) is the set of corruptions and \(f_{model}(t;\Theta)\) is the model-specific scoring function.

Example

>>> import ampligraph.latent_features.loss_functions as lfs
>>> loss = lfs.NLLLoss({'reduction': 'mean'})
>>> isinstance(loss, lfs.NLLLoss)
True

>>> loss = lfs.get('nll')
>>> isinstance(loss, lfs.NLLLoss)
True

Attributes

external_params
name

Methods

__init__([loss_params, verbose])

Initialize the loss..

__init__(loss_params={}, verbose=False)

Initialize the loss..

Parameters:

loss_params (dict) –

Dictionary of hyperparams. No hyperparameters are required for this loss except for “reduction”.

• ”reduction”: (str) - Specifies whether to “sum” or take “mean” of loss per sample w.r.t. corruption (default:”sum”).