Add code to the README

Further 3 million steps. PPO LunarLander-v2 trained agent

Reverting to v2, but trained on 2 million steps. PPO LunarLander-v2 trained agent

Less epochs. PPO LunarLander-v2 trained agent

Higher rollout length, more epochs, lower gamma. PPO LunarLander-v2 trained agent

Update README.md

Upload PPO LunarLander-v2 trained agent