A fork of so-vits-svc with a greatly improved interface. Based on branch 4.0 (v1). No differences in functionality and the models are compatible.
Install this via pip (or your favourite package manager):
pip install -U torch torchaudio --index-url https://download.pytorch.org/whl/cu117
pip install so-vits-svc-fork- Realtime voice conversion
- GUI available
- Unified command-line interface (no need to run Python scripts)
- Ready to use just by installing with
pip. - Automatically download pretrained base model and HuBERT model
- Code completely formatted with black, isort, autoflake etc.
svcg- Realtime (from microphone)
svc --model-path <model-path> source.wav- File
svc vc --model-path <model-path>Colab notebook:
Place your dataset like dataset_raw/{speaker_id}/{wav_file}.wav and run:
svc pre-resample
svc pre-config
svc pre-hubert
svc trainFor more details, run svc -h or svc <subcommand> -h.
svc -h
Usage: svc [OPTIONS] COMMAND [ARGS]...
so-vits-svc allows any folder structure for training data. However, it is
recommended to place the training data in the following structure:
dataset_raw/{speaker_name}/{wav_name}.wav
To train a model, run pre-resample, pre-config, pre-hubert, train. To infer
a model, run infer.
Options:
-h, --help Show this message and exit.
Commands:
clean Clean up files, only useful if you are using the default...
infer Inference
onnx Export model to onnx
pre-config Preprocessing part 2: config
pre-hubert Preprocessing part 3: hubert
pre-resample Preprocessing part 1: resample
train Train model
vc Realtime inference from microphoneThanks goes to these wonderful people (emoji key):
 34j 💻 🤔 📖 |
This project follows the all-contributors specification. Contributions of any kind welcome!