Simple dockerized flask http interface around bel-tts

1. Clone repo
2. docker compose build

1. docker compose up
2. On the first run it will download and caches models and other data from huggingface
3. curl it

curl -X POST http:https://localhost:5001/synthesize \
     -H "Content-Type: application/json" \
     -d '{"text":"ҁбаЦка"}' \
     --output output.wav

4. open output.wav