Fine Tuning

8 problems

• DPO lossMedium
• Generalised Advantage Estimation (GAE)Medium
• GRPO objectiveHard
• Knowledge-Distillation LossMedium
• LoRA forward passEasy
• ORPO Loss (Odds-Ratio Preference Optimisation)Medium
• PPO clipped objectiveMedium
• Reward-Model Margin LossMedium