Understanding Reinforcement Learning with Human Feedback Part 5: Training the Reward Model with Loss Functions