magpie-tts-multilingual

Generate API Key

Pull and Run the NIM

Copy Code
$ docker login nvcr.io
Username: $oauthtoken
Password: <PASTE_API_KEY_HERE>

Refer Supported Models for full list of models.

Copy Code
export NGC_API_KEY=<PASTE_API_KEY_HERE>

docker run -it --rm --name=magpie-tts-multilingual \
    --runtime=nvidia \
    --gpus '"device=0"' \
    --shm-size=8GB \
    -e NGC_API_KEY=$NGC_API_KEY \
    -e NIM_HTTP_API_PORT=9000 \
    -e NIM_GRPC_API_PORT=50051 \
    -p 9000:9000 \
    -p 50051:50051 \
    nvcr.io/nim/nvidia/magpie-tts-multilingual:latest

It may take a up to 30 minutes depending on your network speed, for the container to be ready and start accepting requests from the time the docker container is started.

Open a new terminal and run following command to check if the service is ready to handle inference requests

Copy Code
curl -X 'GET' 'http://localhost:9000/v1/health/ready'

If the service is ready, you get a response similar to the following.

Copy Code
{"ready":true}

Test the NIM

Open a new terminal and run following commands to synthesize audio from text using sample client scripts.

Install the Riva Python client package

Copy Code
sudo apt-get install python3-pip
pip install -U nvidia-riva-client

Download Riva sample clients

Copy Code
git clone https://github.com/nvidia-riva/python-clients.git

Run Text to Speech inference in English (en-US)

Copy Code
python3 python-clients/scripts/tts/talk.py --server 0.0.0.0:50051 \
    --language-code en-US \
    --voice Magpie-Multilingual.EN-US.Sofia \
    --text "Hello, this is a speech synthesizer." \
    --output output_english.wav

Run Text to Speech inference in Spanish (es-US)

Copy Code
python3 python-clients/scripts/tts/talk.py --server 0.0.0.0:50051 \
    --language-code es-US \
    --voice Magpie-Multilingual.ES-US.Diego \
    --text "Vamos al parque los domingos para jugar con nuestros amigos." \
    --output output_spanish.wav

Run Text to Speech inference in French (fr-FR)

Copy Code
python3 python-clients/scripts/tts/talk.py --server 0.0.0.0:50051 \
    --language-code fr-FR \
    --voice Magpie-Multilingual.FR-FR.Louise \
    --text "Je vais au supermarché avec ma mère pour acheter des légumes." \
    --output output_french.wav

On running the above commands, the synthesized audio will be saved to files with corresponding names.

You can list available voices with following command.

Copy Code
python3 python-clients/scripts/tts/talk.py --server 0.0.0.0:50051 \
    --list-voices

For more details on getting started with this NIM, visit the Riva TTS NIM Docs.