“NLP Status Report 2017-8-21”版本间的差异

来自cslt Wiki
跳转至： 导航、 搜索
 
（1位用户的2个中间修订版本未显示）
第19行： 第19行：
 
|-
 
|-
 
|Shipan Ren ||
 
|Shipan Ren ||
−
* organized all the experimental results(our baseline system,Moses,THUMT)  
+
* organized all the experimental results(our baseline system,Moses,THUMT) [http://cslt.riit.tsinghua.edu.cn/mediawiki/images/8/89/Nmt_baseline.xlsx]
 
* trained and tested translation models（Toolkit:THUMT ）
 
* trained and tested translation models（Toolkit:THUMT ）
 
* compared with our system
 
* compared with our system
第45行： 第45行：
 
*3.data used Shiji and Zizhitongjian(43,0000 pairs), we can get BLEU about 9.
 
*3.data used Shiji and Zizhitongjian(43,0000 pairs), we can get BLEU about 9.
 
*4.data used Shiji and Zizhitongjian(43,0000 pairs), and split the ancient language text one character by one, we can get BLEU 11.11 at most.
 
*4.data used Shiji and Zizhitongjian(43,0000 pairs), and split the ancient language text one character by one, we can get BLEU 11.11 at most.
−
*The main factors now is the data(including pairs of sentence、the quality——cause the modern language text include context information.
+
*The main factors now is the data(pairs of sentence/the quality——the modern language text includes context information).
 
||   
 
||   
−
*plan to read source code of seq2seq model;
+
*plan to read source code of seq2seq model and learn tensorflow;
 
*plan to read a paper named Automatic Long Sentence Segmentation for NMT
 
*plan to read a paper named Automatic Long Sentence Segmentation for NMT
 
|-
 
|-
 
|}
 
|}

2017年8月28日 (一) 07:44的最后版本

Date People Last Week This Week
2017/8/14 Jiyuan Zhang
  • done some work about code refactoring for poem system
  • plan to complete code refactoring for poem system
Aodong LI
Shiyue Zhang
Shipan Ren
  • organized all the experimental results(our baseline system,Moses,THUMT) [1]
  • trained and tested translation models（Toolkit:THUMT ）
  • compared with our system
  • prepare to release the baseline system（tensorflow1.0 version）
Jiayu Guo
  • process data and run model;
  • test results.

checkpoint-100000 translation model BLEU： 11.11

  • source:在秦者名错，与张仪争论,於是惠王使错将伐蜀，遂拔，因而守之。
  • target:在秦国的名叫司马错，曾与张仪发生争论，秦惠王采纳了他的意见，于是司马错率军攻蜀国，攻取后，又让他做了蜀地郡守。
  • trans：当时秦国的人都很欣赏他的建议，与张仪一起商议，所以吴王派使者率军攻打蜀地，一举攻，接着又下令守城 。
  • source:神大用则竭，形大劳则敝，形神离则死 。
  • target:精神过度使用就会衰竭，形体过度劳累就会疲惫，神形分离就会死亡。
  • trans: 精神过度就可衰竭,身体过度劳累就会疲惫，地形也就会死。
  • source:今天子接千岁之统，封泰山，而余不得从行，是命也夫，命也夫！
  • target:现天子继承汉朝千年一统的大业，在泰山举行封禅典礼而我不能随行，这是命啊，是命啊！
  • trans: 现在天子可以继承帝位的成就爵位，爵位至泰山，而我却未能执行先帝的命运。
  • 1.data used Zizhitongjian only(6,000 pairs), we can get BLEU 6 at most.
  • 2.data used Zizhitongjian only(12,000 pairs), we can get BLEU 7 at most.
  • 3.data used Shiji and Zizhitongjian(43,0000 pairs), we can get BLEU about 9.
  • 4.data used Shiji and Zizhitongjian(43,0000 pairs), and split the ancient language text one character by one, we can get BLEU 11.11 at most.
  • The main factors now is the data(pairs of sentence/the quality——the modern language text includes context information).
  • plan to read source code of seq2seq model and learn tensorflow;
  • plan to read a paper named Automatic Long Sentence Segmentation for NMT
取自“http://index.cslt.org/mediawiki/index.php?title=NLP_Status_Report_2017-8-21&oldid=28461”