Search for a command to run...
TADA: A Generative Framework for Speech Modeling via Text-Acoustic Dual Alignment