Original repo: StyleTTS2

Paper: https://arxiv.org/abs/2306.07691

Environment setup

1. Python >= 3.7
2. Clone this repository:

git clone https://github.com/5Hyeons/STTS2_server.git
cd STTS2_server

3. If you don’t have PyTorch installed, follow the instructions here to install the appropriate versions of torch and torchaudio.

4. Install python requirements:

pip install -r requirements.txt

• Note: You must have cuDNN installed to use onnxruntime-gpu.

5. Download the models here and place them in the STTS2_server/Models directory.

Run server

• To run the server using a .pth model:

python server.py

• To Run the server using an ONNX model:

python server_onnx.py