@ -3,15 +3,17 @@ This example contains code used to train a [parallel wavegan](http://arxiv.org/a
## Dataset
## Dataset
### Download and Extract
### Download and Extract
Download CSMSC from it's [official website](https://test.data-baker.com/data/index/TNtts/) and extract it to `~/datasets`. Then the dataset is in the directory `~/datasets/BZNSYP`.
Download CSMSC from it's [official website](https://test.data-baker.com/data/index/TNtts/) and extract it to `~/datasets`. Then the dataset is in the directory `~/datasets/BZNSYP`.
datasets/BZNSYP should have three folders:
After processing the data, the ``BZNSYP`` directory will look like this:
└─ Wave
```text
└─ .wav files (audio speech)
BZNSYP
└─ PhoneLabeling
├── Wave
└─ .interval files (alignment between phoneme and duration)
│ └─ *.wav files (audio speech)
└─ ProsodyLabeling
├── PhoneLabeling
└─ 000001-010000.txt (text with prosodic by pinyin)
│ └─ *.interval files (alignment between phoneme and duration)
Still we only use .wav files in training.
└── ProsodyLabeling
└─ 000001-010000.txt (text with prosodic by pinyin)
```
### Get MFA Result and Extract
### Get MFA Result and Extract
We use [MFA](https://github.com/MontrealCorpusTools/Montreal-Forced-Aligner) results to cut silence at the edge of audio.
We use [MFA](https://github.com/MontrealCorpusTools/Montreal-Forced-Aligner) results to cut silence at the edge of audio.