Simple Fastapi example to make inference with tts matxa multispeaker model in onnx format.
Steps:
- Clone the repo.
git clone https://github.com/langtech-bsc/minimal-tts-multispeaker-api - Build and run the container.
cd minimal-tts-multispeaker-api
docker build -t minimal-tts-multispeaker-api .
docker run -p 8000:8000 -t minimal-tts-multispeaker-api
- Test with a simple request.
curl -X POST http://0.0.0.0:8000/api/tts -H "Content-Type: application/json" -d '{"text":"Bon dia.","voice":20,"type":"text"}' | play -t wav -