http://cslt.org/mediawiki/index.php?action=history&feed=atom&title=Huilian-cslt-week-2016-1-7Huilian-cslt-week-2016-1-7 - 版本历史2026-04-14T07:08:06Z本wiki的该页面的版本历史MediaWiki 1.23.3http://cslt.org/mediawiki/index.php?title=Huilian-cslt-week-2016-1-7&diff=18535&oldid=prevZxw：以“==Environment== ==RNN Poem Processing== * attention training from Renmin newspaper done --Qixin Wang done * change corpus done * contact specialists of Song Pe...”为内容创建页面2016-01-07T02:02:27Z<p>以“==Environment== ==RNN Poem Processing== * attention training from Renmin newspaper done --Qixin Wang done * change corpus done * contact specialists of Song Pe...”为内容创建页面</p>
<p><b>新页面</b></p><div>==Environment==<br />
<br />
<br />
==RNN Poem Processing==<br />
* attention training from Renmin newspaper done --Qixin Wang done <br />
* change corpus done <br />
* contact specialists of Song Peams done <br />
* write a patent done<br />
* submit a patent --Xinkai(1.15)<br />
* demo release(unified model) done<br />
* add to language model information<br />
* crowdsourcing platform check <br />
<br />
==RNN Writing Processing==<br />
* write code done -- Chao Xing<br />
* generate Xing style using variational RNN <br />
:* use mnist dataset done<br />
:* train using chinese dataset done <br />
* train reference to the paper of BAYESIAN PROBABILISTIC PROGRAM FOR IMAGE GENERATION <br />
<br />
==RNN QA==<br />
* rnn rescore<br />
:* rnn rescore should be used as a factor for learning to rank <br />
* similar question --Bingdong<br />
** Huilan test -- Bingdong<br />
*** train corpus of word vector--hulian<br />
*** add traditional method of learning to rank--Tianyi<br />
* memory machine(about three months)--Tianyi<br />
<br />
==Couplet==<br />
* attention training done --Tianyi Luo<br />
* focused attending training<br />
* revise training corpus</div>Zxw