Huizhuo Angela Yuan

The official implementation of Self-Play Fine-Tuning (SPIN)

The official implementation of MARS: Unleashing the Power of Variance Reduction for Training Large Models

The official implementation of Self-Play Preference Optimization (SPPO)