“Huilian-cslt-week-2016-3-17”版本间的差异

来自cslt Wiki
跳转至: 导航搜索
(以“==Environment== ==RNN Poem Processing== * paper release --Qixin * update a new version(paper release first) --Qixin ==RNN Writing Processing== * neural model le...”为内容创建页面)
 
Zxw讨论 | 贡献
 
第1行: 第1行:
 
==Environment==
 
==Environment==
 +
* buy a new server --du
  
 
==RNN Poem Processing==   
 
==RNN Poem Processing==   
 
* paper release --Qixin  
 
* paper release --Qixin  
 
* update a new version(paper release first) --Qixin
 
* update a new version(paper release first) --Qixin
 +
  
 
==RNN Writing Processing==
 
==RNN Writing Processing==
第13行: 第15行:
 
* attention based QA (Solution for template expansion)--Dongxu
 
* attention based QA (Solution for template expansion)--Dongxu
 
:* compute probability done  
 
:* compute probability done  
:* capability generalization
 
 
:* POS similar pair  
 
:* POS similar pair  
:* LSTM attention
+
:* LSTM attention done
 +
:* evaluation unseen candidate
 +
:* collect chatting data --du
 +
::* train chatting model using tensor flow(--Dongxu) or theano(--Qixin)
 
* attention and memory QA--Tianyi
 
* attention and memory QA--Tianyi
 
* Dongxu Investigation first
 
* Dongxu Investigation first

2016年3月17日 (四) 02:28的最后版本

Environment

  • buy a new server --du

RNN Poem Processing

  • paper release --Qixin
  • update a new version(paper release first) --Qixin


RNN Writing Processing

  • neural model learnig transform --Xingchao
  • add attention
  • BPL --Yanxu

RNN QA

  • attention based QA (Solution for template expansion)--Dongxu
  • compute probability done
  • POS similar pair
  • LSTM attention done
  • evaluation unseen candidate
  • collect chatting data --du
  • train chatting model using tensor flow(--Dongxu) or theano(--Qixin)
  • attention and memory QA--Tianyi
  • Dongxu Investigation first
  • rnn rescore--Bingdong
  • rnn rescore should be used as a factor for learning to rank
  • query expansion--Bingdong

Couplet

  • bug fits--Tianyi
  • test --Yiqiao (3.17)