Supervised Fine-Tuning as Inverse Reinforcement Learning

Re-assign community