|
--- |
|
tags: |
|
- espnet |
|
- audio |
|
- text-to-speech |
|
language: en |
|
datasets: |
|
- ljspeech |
|
license: cc-by-4.0 |
|
--- |
|
|
|
TTS model trained with Montreal Forced Aligner. |
|
|
|
To replicate or continue training from the given checkpoint, download [LJSpeech](https://keithito.com/LJ-Speech-Dataset/), |
|
install [MFA](https://github.com/MontrealCorpusTools/Montreal-Forced-Aligner) |
|
and follow the steps [here](https://github.com/espnet/espnet/blob/master/egs2/TEMPLATE/tts1/README.md). |
|
I recommend downloading the [pretrained MFA models](https://mfa-models.readthedocs.io/en/latest/) and running `mfa.sh` with `--train false`. |
|
This would help to save time, but one disadvantage is you have to use the MFA g2p for inference (which has a non-standard phoneme set). |