continuous output for nmt

afd5a5ac · zengxin · afd5a5ac · afd5a5ac · afd5a5ac · afd5a5ac
Commit afd5a5ac authored Dec 10, 2020 by zengxin
--- a/Continuous Output for NMT/A Margin-based Loss with Synthetic Negative Samples for continuous output machine translation.pdf
+++ b/Continuous Output for NMT/A Margin-based Loss with Synthetic Negative Samples for continuous output machine translation.pdf
--- a/Continuous Output for NMT/Beyond BLEU Training Neural Machine Translation with Semantic Similarity.ACL2019.pdf
+++ b/Continuous Output for NMT/Beyond BLEU Training Neural Machine Translation with Semantic Similarity.ACL2019.pdf
--- a/Continuous Output for NMT/Efficient Contextual Representation Learning with continuous outputs.pdf
+++ b/Continuous Output for NMT/Efficient Contextual Representation Learning with continuous outputs.pdf
--- a/Continuous Output for NMT/ReWE RegressingWord Embeddings for Regularization of Neural Machine Translation Systems.NAACL2019.pdf
+++ b/Continuous Output for NMT/ReWE RegressingWord Embeddings for Regularization of Neural Machine Translation Systems.NAACL2019.pdf
--- a/Continuous Output for NMT/VON MISES-FISHER LOSS FOR TRAINING SEQUENCE.pdf
+++ b/Continuous Output for NMT/VON MISES-FISHER LOSS FOR TRAINING SEQUENCE.pdf
--- a/Continuous Output for NMT/continuous output fot nmt.md
+++ b/Continuous Output for NMT/continuous output fot nmt.md
+* Beyond BLEU: Training Neural Machine Translation with Semantic Similarity，ACL 2019
+  * BLEU 并不完全可信，他会给语义相似但词汇不同的句子以惩罚。本文介绍一种新的评价指标SIMILE。
+  * SIMILE是一种对句子相似度进行“连续”度量的指标，借鉴了专注于领域不可知的语义相似度的度量工作。