# preload.py from TTS.api import TTS # Pre-download TTS model so it’s ready at runtime tts = TTS(model_name="tts_models/en/ljspeech/glow-tts")