論文分享 -- > NLP -- > Neural machine Translation of Rare Words with Subword Units
本次總結的是一篇16年的關於NLP中分詞操做的論文，論文連接Subword，參考的實現代碼subword-nmt，許多論文方法（例如BERT等）都將該方法應用到分詞處理上，相對於word-level和character-level，該方法取得了不錯的效果。git 動機和創新點 機器翻譯中，一般使用固定大小的詞表，而在實際翻譯場景中，應當是open-vocabulary。這就使得翻譯數據集中的稀有詞
相關文章
- 1. Information from parts of words: Subword Models
- 2. 【CS224n】Neural Machine Translation with Seq2Seq
- 3. 論文分享 -- > NLP -- > Language Models are Unsupervised Multitask Learners
- 4. 【CS224n】Lecture 12: Information from parts of words: Subword Models
- 5. [論文閱讀筆記] Massive Exploration of Neural Machine Translation Architectures
- 6. 論文解讀：Multimodal Machine Translation with Embedding Prediction
- 7. Neural machine translation by jointly learning to align and translate論文閱讀筆記
- 8. 論文分享-->Attention-over-Attention Neural Networks for Reading Comprehension
- 9. 論文閱讀——Unsupervised Neural Machine Translation with Indirect Supervision
- 10. 【cs224n-10】Information from parts of words (Subword Models)
- 更多相關文章...
- • Docker Machine - Docker教程
- • CAP理論是什麼？ - NoSQL教程
- • Git五分鐘教程
- • 算法總結-二分查找法
相關標籤/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
歡迎關注本站公眾號,獲取更多信息
相關文章
- 1. Information from parts of words: Subword Models
- 2. 【CS224n】Neural Machine Translation with Seq2Seq
- 3. 論文分享 -- > NLP -- > Language Models are Unsupervised Multitask Learners
- 4. 【CS224n】Lecture 12: Information from parts of words: Subword Models
- 5. [論文閱讀筆記] Massive Exploration of Neural Machine Translation Architectures
- 6. 論文解讀：Multimodal Machine Translation with Embedding Prediction
- 7. Neural machine translation by jointly learning to align and translate論文閱讀筆記
- 8. 論文分享-->Attention-over-Attention Neural Networks for Reading Comprehension
- 9. 論文閱讀——Unsupervised Neural Machine Translation with Indirect Supervision
- 10. 【cs224n-10】Information from parts of words (Subword Models)