main
[Coati] Train DPO using PP (#6054)
fix style
fix style, add kto data sample
Add GRPO and Support RLVR for PPO (#6186)
remove real data path