A latent-variable non-autoregressive translation model with deterministic delta-posterior inference matches autoregressive quality within 2 BLEU points while decoding 12.5x faster.
Title resolution pending
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
-
Latent-Variable Non-Autoregressive Neural Machine Translation with Deterministic Inference Using a Delta Posterior
A latent-variable non-autoregressive translation model with deterministic delta-posterior inference matches autoregressive quality within 2 BLEU points while decoding 12.5x faster.