KRAMPUS KART: PPO
How are cars doing?
Cars: 0/0 | Leader: 0%
Avg Reward: 0 | Best: 0
How is learning going?
PPO updates: 0
Initializing...
Actor: —
Critic: —
What is PPO?
Actor = the policy (decides steering)
Critic = predicts future reward
Advantage = "was this action good?"
Car Colors = Episode Reward
Separate Networks (as paper recommends)
Line colors: red=negative, blue=positive | Each trained independently!
Avg Reward