lander-ppo_rl

🚀 PPO Agent for LunarLander-v2

This is a trained PPO agent that learned to land a spacecraft on the moon!

Model Description

• Algorithm: Proximal Policy Optimization (PPO)
• Environment: LunarLander-v2
• Framework: Stable-Baselines3
• Training Steps: 100,000 - 500,000 steps

Performance

• Success Rate: 90%+ successful landings
• Average Reward: 200+ (successful landing threshold)
• Best Performance: 265+ reward

Usage

from stable_baselines3 import PPO
import gymnasium as gym

# Load the trained model
model = PPO.load("lunar_lander_ppo_model")

# Create environment
env = gym.make('LunarLander-v2', render_mode='human')

# Test the agent
obs, _ = env.reset()
for _ in range(1000):
    action, _ = model.predict(obs, deterministic=True)
    obs, reward, terminated, truncated, info = env.step(action)
    if terminated or truncated:
        obs, _ = env.reset()

env.close()

Training Details

The agent was trained using PPO with the following hyperparameters:

• Learning rate: 0.0003
• Batch size: 64
• Number of environments: 4
• Gamma: 0.999

Results

The agent successfully learned to:

• Control spacecraft thrust
• Navigate to landing pad
• Execute gentle landings
• Conserve fuel efficiently

Watch it land on the moon! 🌙