“Zhiyuan Tang”版本间的差异
来自cslt Wiki
（以“ Last week: 1. With Mengyuan, conducted experiments for small model with big data, initial results [http://192.168.0.51:5555/cgi-bin/cvss/cvss_request.pl?account=t...”为内容创建页面） |
|||
| 第11行： | 第11行： | ||
1. transfer learning(dark knowledge) for nnet3; | 1. transfer learning(dark knowledge) for nnet3; | ||
| − | 2. speech & speaker recognition | + | 2. speech & speaker recognition joint training. |
2016年2月22日 (一) 00:19的版本
Last week:
1. With Mengyuan, conducted experiments for small model with big data, initial results here;
2. got the conlusion that mpe improves chain, results here.
This week:
1. transfer learning(dark knowledge) for nnet3;
2. speech & speaker recognition joint training.
