Data-efficient Targeted Token-level Preference Optimization for LLM-based Text-to-Speech

Open in new window