Updated Model v01 Tacotron2, DDC PWGAN (markdown)

Thorsten Müller 2020-09-14 20:15:41 +02:00
parent 0c0cada115
commit 0d768a1e68

@ -1,14 +1,23 @@
Based on "thorsten" voice contribution
Based on "thorsten" voice dataset v02 (normalized)
# Release Candidate 1 - "Early bird"
This early bird rc was published on september 2020 and is meant for first experiments.
# Release Candidate 1 - "early bird"
This early bird rc was published on september, 14th 2020 and is meant for first experiments.
All files (configs and checkpoints on google drive.
## Content of this rc
* [x] Train tacotron2 model based on version 02 of dataset (normalized) with ddc (thanks to Olaf)
* [x] Train pwgan vocoder model
* [ ] number 2
* [x] train tacotron2 model based on version 02 of dataset (normalized) with ddc up to 360k steps (thanks to Olaf)
> Which checkpoints should be provided?
* [x] train pwgan vocoder model (2,75 mio. steps) thanks to erogol
> Which checkpoints should be provided?
* [x] provide audio samples
* [x] jupyter notebook for testing
* [x] usage of german phoneme cleaner (thanks to repodiac)
* [x] usage of english loan words
* [x] use compute_statistics output in vocoder training
* [ ] tbd
## Audio samples
* soundcloud link
## How to use
* Link to Olafs notebook