PPO and GRPO in LLMs

An explainer comparing Proximal Policy Optimization (PPO) and Group Relative Policy Optimization (GRPO) in reinforcement learning for language models.