Pseudo self attention, which injects learned encoder outputs into the self-attention of a pretrained language model, consistently outperforms prior encoder-agnostic adaptation methods on four conditional generation tasks.
Simple Fusion: Return of the Language Model
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
Neural Machine Translation (NMT) typically leverages monolingual data in training through backtranslation. We investigate an alternative simple method to use monolingual data for NMT training: We combine the scores of a pre-trained and fixed language model (LM) with the scores of a translation model (TM) while the TM is trained from scratch. To achieve that, we train the translation model to predict the residual probability of the training data added to the prediction of the LM. This enables the TM to focus its capacity on modeling the source sentence since it can rely on the LM for fluency. We show that our method outperforms previous approaches to integrate LMs into NMT while the architecture is simpler as it does not require gating networks to balance TM and LM. We observe gains of between +0.24 and +2.36 BLEU on all four test sets (English-Turkish, Turkish-English, Estonian-English, Xhosa-English) on top of ensembles without LM. We compare our method with alternative ways to utilize monolingual data such as backtranslation, shallow fusion, and cold fusion.
fields
cs.CL 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Encoder-Agnostic Adaptation for Conditional Language Generation
Pseudo self attention, which injects learned encoder outputs into the self-attention of a pretrained language model, consistently outperforms prior encoder-agnostic adaptation methods on four conditional generation tasks.