“Mengyuan Zhao 2015-12-14”版本间的差异
来自cslt Wiki
| 第2行： | 第2行： | ||
*AM training on big dataset --mengyuan | *AM training on big dataset --mengyuan | ||
:* Test nnet3-MPE code from Xuewei on sinovoice 120h 16k dataset. Didn't observe performance improvement. There are still some bugs in the code. | :* Test nnet3-MPE code from Xuewei on sinovoice 120h 16k dataset. Didn't observe performance improvement. There are still some bugs in the code. | ||
| − | :* Run nnet3-ctc training on sinovoice 120h 16k dataset. Result looks ok, but worse normal model. | + | :* Run nnet3-ctc training on sinovoice 120h 16k dataset. Result looks ok, but worse than normal model. |
:* Start nnet3-ctc training on sinovoice 4000h 8k dataset. | :* Start nnet3-ctc training on sinovoice 4000h 8k dataset. | ||
2015年12月13日 (日) 12:44的最后版本
RNN AM
- AM training on big dataset --mengyuan
- Test nnet3-MPE code from Xuewei on sinovoice 120h 16k dataset. Didn't observe performance improvement. There are still some bugs in the code.
- Run nnet3-ctc training on sinovoice 120h 16k dataset. Result looks ok, but worse than normal model.
- Start nnet3-ctc training on sinovoice 4000h 8k dataset.
