A controllable dual skew divergence loss, in which a PI controller adjusts the balance between forward and reverse KL terms, improves BLEU by 0.2 to 1.1 points over cross-entropy baselines on WMT14 En-De and En-Fr.
Sequence to sequence learning with neural networks,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Controllable Dual Skew Divergence Loss for Neural Machine Translation
A controllable dual skew divergence loss, in which a PI controller adjusts the balance between forward and reverse KL terms, improves BLEU by 0.2 to 1.1 points over cross-entropy baselines on WMT14 En-De and En-Fr.