speechLLMs Forked from https://github.com/Hao-Shi-SBINT/Multi-talker-ASR-with-LLMs Include speech emotion recognition in a unified structure.