Commit Graph

  • c25c62b8f9 refine audio_data_utils.py dangqingqing 2017-06-12 19:06:55 +0800
  • bb34e90398 correct typos in annotations Yibing Liu 2017-06-12 17:20:22 +0800
  • ae83a25aff add beam search decoder using multiprocesses Yibing Liu 2017-06-12 17:13:48 +0800
  • 453f038df9 optimize the efficiency of beam search Yibing Liu 2017-06-08 16:05:40 +0800
  • 9fda521ee3 improve external scorer Yibing Liu 2017-06-07 19:24:04 +0800
  • b046e651e7 tiny modify to pass CI Yibing Liu 2017-06-07 17:43:12 +0800
  • 44efbed798 rename variables in decoder Yibing Liu 2017-06-07 16:59:11 +0800
  • 21ff590e6d modify language model scoring Yibing Liu 2017-06-07 14:57:04 +0800
  • ac370eca85 add annotations Yibing Liu 2017-06-07 09:06:58 +0800
  • 51f35a5372 code clean & add external scorer Yibing Liu 2017-06-07 02:21:35 +0800
  • dedbfb2654 enable ctc beam search decoder Yibing Liu 2017-06-05 08:48:54 +0800
  • cfe9d22866 update annotations Yibing Liu 2017-06-04 19:19:36 +0800
  • c943ca79ac mv ctc_beam_search_decoder into deep_speech_2/ Yibing Liu 2017-06-04 19:15:09 +0800
  • e016778e20 Merge pull request #78 from xinghai-sun/ds2_libri Xinghai Sun 2017-06-09 11:52:10 +0800
  • 06e9f71389 Remove manifest's line number check from librispeech.py and update README.md. Xinghai Sun 2017-06-08 22:20:11 +0800
  • 3f63e069e0 Fix typos and follow comments. yangyaming 2017-06-08 21:35:17 +0800
  • cc2a4d4e3d Add error rate calculation script. yangyaming 2017-06-08 17:18:38 +0800
  • f49eab5fec Change assert to exception raising. yangyaming 2017-06-07 19:11:21 +0800
  • d3eeb7fd76 Refine librispeech.py for DeepSpeech2. Xinghai Sun 2017-06-07 17:44:11 +0800
  • 7c85e0fdb5 Support variable input batch and sortagrad. dangqingqing 2017-06-07 16:37:13 +0800
  • d2e467385d Add loading model function for train.py. yangyaming 2017-06-05 21:00:15 +0800
  • 730d5c4dd3 Update DS2 README.md and fix bug in librispeech.py Xinghai Sun 2017-06-03 15:18:33 +0800
  • 2a83486500 Refactor decoder interfaces and add ./data directory. Xinghai Sun 2017-06-03 14:52:02 +0800
  • 8313895e85 1. Fix incorrect decoder result printing. 2. Fix incorrect batch-norm usage in RNN. 3. Fix overlapping train/dev/test manfests. 4. Update README.md and requirements.txt. 5. Expose more arguments to users in argparser. 6. Update all other details. Xinghai Sun 2017-06-02 21:07:58 +0800
  • e6a349992b Refactor data utils into a class and add feature normalization. Xinghai Sun 2017-05-30 20:34:03 +0800
  • 9c3cd3c704 Update some parameters and comments. Xinghai Sun 2017-05-26 17:47:24 +0800
  • 0babc5c4d7 Add function docs. Xinghai Sun 2017-05-25 23:36:06 +0800
  • 70a343a499 Add infererence and add SortaGrad for only first pass. Xinghai Sun 2017-05-25 16:10:23 +0800
  • 3fc94427db Add librispeech dataset, audio data provider and simplfied DeepSpeech2 model configuration. Bug exists when run training. Xinghai Sun 2017-05-25 01:17:18 +0800
  • d59b8ca97e Add deep_speech_2 folder. Xinghai Sun 2017-05-22 21:00:37 +0800