(Suggestion) Python == 3.7
git clone https://github.com/CjangCjengh/vits.git
- Fill "text_cleaners" in config.json
- Edit text/symbols.py
- Remove unnecessary imports from text/cleaners.py
pip install -r requirements.txt
"n_speakers" should be 0 in config.json
path/to/XXX.wav|transcript
- Example
dataset/001.wav|こんにちは。
Speaker id should start from 0
path/to/XXX.wav|speaker id|transcript
- Example
dataset/001.wav|0|こんにちは。
If you have done this, set "cleaned_text" to true in config.json
# Single speaker
python preprocess.py --text_index 1 --filelists path/to/filelist_train.txt path/to/filelist_val.txt
# Mutiple speakers
python preprocess.py --text_index 2 --filelists path/to/filelist_train.txt path/to/filelist_val.txt
cd monotonic_align
python setup.py build_ext --inplace
cd ..
# Single speaker
python train.py -c <config> -m <folder>
# Mutiple speakers
python train_ms.py -c <config> -m <folder>
See inference.ipynb
See MoeGoe
docker run -itd --gpus all --name "Container name" -e NVIDIA_DRIVER_CAPABILITIES=compute,utility -e NVIDIA_VISIBLE_DEVICES=all "Image name"
python train_ms.py -c configs/yosuga.json -m yosuga -o /root/tf-logs