/speech-trident

Awesome speech/audio LLMs, representation learning, and codec models

Watchers