Commit Graph

26 Commits (d49a938a71b0e3a32e6dee0950c4e50cc75faa65)

Author SHA1 Message Date
Keith Ito d49a938a71 Update documentation 2017-09-04 21:20:25 -07:00
Keith Ito 479976b6c5 Simplify text processing, make it easier to adapt to non-English data. 2017-09-04 20:04:59 -07:00
Keith Ito 0894cd9bf2 Make power a hyperparameter 2017-09-03 12:36:09 -07:00
Keith Ito 6a793ba8c4 Update model download instructions
Some versions of tar require the "j" flag to decompress
2017-08-22 17:42:26 -04:00
Keith Ito ae82a7c1ae Merge pull request #21 from keithito/data-instructions
Add documentation on preprocessing training data.
2017-08-14 11:04:09 -04:00
Keith Ito 3c211e7a19 Add documentation on preprocessing training data. 2017-08-14 11:02:19 -04:00
Keith Ito 516ff9db55 Update README to point at latest model 2017-07-25 09:23:30 -04:00
Keith Ito b53447196d Add note about gradient clipping fix 2017-07-24 11:20:29 -04:00
Keith Ito aca83c8298 Merge pull request #7 from candlewill/patch-3
use clipped_gradients not the original gradients
2017-07-24 11:17:50 -04:00
Yunchao He 99518173c0 use clipped_gradients not the original gradients
To my understanding, it's better to use the `clipped_gradients` not the original `gradients`. Is that right?
2017-07-24 17:32:07 +08:00
Keith Ito cdf389bf0b Merge pull request #4 from candlewill/patch-2
specify encoding method
2017-07-17 13:29:16 -04:00
Keith Ito 7ebe9ba6bc Merge pull request #3 from candlewill/patch-1
specify encoding method
2017-07-17 13:28:59 -04:00
Yunchao He 1cfac99798 specify encoding method
specify encoding method
2017-07-17 22:53:06 +08:00
Yunchao He 9081c76b9a specify encoding
specify encoding=utf-8
2017-07-17 22:44:06 +08:00
Keith Ito cdcca53f54 Fix indentation in README 2017-07-15 18:42:46 -04:00
Keith Ito b675f0efa8 Update README to reference 500k samples and model. 2017-07-15 18:40:56 -04:00
Keith Ito 8b514cfffb Update README 2017-07-12 08:36:29 -07:00
Keith Ito 5d332df86b Merge pull request #1 from keithito/audio-quality
Small tweaks to improve audio quality
2017-07-12 08:35:34 -07:00
Keith Ito b01c2382cc Update README 2017-07-12 08:24:03 -07:00
Keith Ito d8910bcb84 Scale audio to use full 16-bit range 2017-07-12 08:21:40 -07:00
Keith Ito cc9ee0073b Raising magnitude to higher power prior to GL seems to reduce artifacts. 2017-07-12 08:16:23 -07:00
Keith Ito 3ad67c9d02 Spell miscellaneous right 2017-07-08 19:16:39 -04:00
Keith Ito 0868767148 Add a note about training with cmudict 2017-07-08 18:53:14 -04:00
Keith Ito 5a51708b7b Add a header for the sample output section 2017-07-08 13:15:43 -04:00
Keith Ito bc706654ad Initial commit 2017-07-08 13:08:26 -04:00
Keith Ito 3ab028c4e6 Initial commit 2017-07-08 13:03:32 -04:00