Reinforcement Learning with Human Feedback (RLHF) – How to train and fine-tune Transformer Models