so-vits-svcを試す

git clone git@github.com:svc-develop-team/so-vits-svc.git

python -m venv venv
.\venv\Scripts\Activate.ps1

pip install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu117

pip install SoundFile
pip install sounddevice
pip install requests==2.28.1
pip install onnx onnxsim onnxoptimizer
pip install scipy==1.9.3
pip install Flask==2.1.2 Flask_Cors==3.0.10
pip install playsound==1.3.0 PyAudio==0.2.12
pip install pydub==0.25.1
pip install pyworld
pip install tqdm==4.63.0
pip install scikit-maad
pip install praat-parselmouth
pip install gradio==3.9
pip install starlette==0.20.4
pip install fairseq==0.12.2

pip install tensorboard

pip install librosa

error: Microsoft Visual C++ 14.0 or greater is required. Get it with "Microsoft C++ Build Tools": https://visualstudio.microsoft.com/visual-cpp-build-tools/

wget -P hubert/ http://obs.cstcloud.cn/share/obs/sankagenkeshi/checkpoint_best_legacy_500.pt

dataset_raw
├───speaker0
│   ├───xxx1-xxx1.wav
│   ├───...
│   └───Lxx-0xx8.wav
└───speaker1
    ├───xx2-0xxx2.wav
    ├───...
    └───xxx7-xxx007.wav

python resample.py

python preprocess_flist_config.py

python preprocess_hubert_f0.py

{
  "train": {
    "log_interval": 200,
    "eval_interval": 800,
    "seed": 1234,
    "epochs": 1000,
    "learning_rate": 0.0001,
    "betas": [
      0.8,
      0.99
    ],
    "eps": 1e-09,
    "batch_size": 6,

python train.py -c configs/config.json -m 44k

sudo apt install ffmpeg
ffmpeg -i "input.mp3" -vn -ac 2 -ar 44100 -acodec pcm_s16le -f wav "output.wav"

python inference_main.py -m "logs/44k/G_1000.pth" -c "configs/config.json" -n "output.wav" -t 0 -s "tsukuyomi" -a