Reinforcement Learning for Wordle