Improving Robustness of LLM-based Speech Synthesis by Learning Monotonic Alignment

Re-assign community