Upload PPO LunarLander-v2 trained agent third attempt, longer training

Upload PPO LunarLander-v2 trained agent second attempt, longer training + new params

First version, standard with 1000000 training steps

initial commit