Language zh is not supported by the espeakbackend name

[JRMeyer]

>>> Margarita
[October 5, 2019, 12:34am]

Hi,

I'm trying to train a new Chinese tts dataset using Tacotron2 on
Dev branch

If I set 'use_phonme'=true, and 'phoneme_language'=zh, then following
errors occur: slash

19-34-40.png.filename}[1168×435 73.1

19-34-52.png.filename}[1214×491 107

The version is eSpeak text-to-speech: 1.48.03 and phonemizer-1.1 slash
available backends: festival-2.5.0, espeak-ng-1.49.2, segments-2.1.0

It's confusing again that espeak with language 'zh' worked well on
master branch training tacotron

Anyone been through this before? Thanks!

[This is an archived TTS discussion thread from discourse.mozilla.org/t/language-zh-is-not-supported-by-the-espeakbackend-name]

[JRMeyer]

>>> nmstoker
[October 5, 2019, 1:33am]

Do you manage to make Espeak output speech on the command line using zh?
If you can't get that working then phonemiser (and in turn TTS) is also
likely to struggle.

I've no experience with training for Chinese but I wonder whether you
might be better off with espeak-ng, and if you use that, then I believe
the language code is slightly different: cmn

github.com

#### espeak-ng/espeak-ng/blob/master/docs/languages.md

# Languages

The languages in espeak-ng are grouped by their
ISO 639-5 language
family code. They are identified by their
BCP 47 identifier. For several accents
and dialects,
private-use extensions
have been used.

The 107 supported languages and accents are:

| Family Code | Identifier | Language Family | Language | Accent/Dialect |
|-------------|-------------------|-----------------------|-----------------------------|------------------------|
| gmw | af | West Germanic | Afrikaans | |
| ine | sq | Indo-European | Albanian | |
| sem | am | Semitic | Amharic | |
| sem | ar | Semitic | Arabic ^{slash [1 slash ]} | |
| roa | an | Romance | Aragonese | |
| ine | hy | Indo-European | Armenian | East Armenian |

This file has been truncated. show
original

[Archived Post]