Self-Play Fine-Tuning Converts Weak Language Models to Strong Language Models

Chen, Zixiang, Deng, Yihe, Yuan, Huizhuo · arXiv · 2024