ppo-LunarLander-v2

PPO Agent playing LunarLander-v2 🚀🌕

This is a trained Deep Reinforcement Learning agent playing the LunarLander-v2 environment. The agent was trained using the Proximal Policy Optimization (PPO) algorithm from the stable-baselines3 library.

🎯 Project Overview

The objective of this project was to train an autonomous agent to safely navigate a lunar module to its landing pad. The agent must manage fuel efficiency, gravity, and precise steering in a continuous state space.

This model successfully "solved" the environment by achieving a consistent score well over the 200-point threshold.

📈 Evaluation Results

• Mean Reward: 238.00 +/- 14.39
• Status: Environment Solved

🧠 Model Details

• Algorithm: PPO (Proximal Policy Optimization)
• Architecture: Actor-Critic Deep Neural Network (MLP)
• Library: stable-baselines3
• Environment: gymnasium (LunarLander-v2)

💻 Usage (with Stable-baselines3)

You can easily load and watch this agent play using the huggingface_sb3 library!

import gymnasium as gym
from stable_baselines3 import PPO
from huggingface_sb3 import load_from_hub

# 1. Download the model from the Hugging Face Hub
repo_id = "imbibek8366/ppo-LunarLander-v2"
filename = "ppo-LunarLander-v2.zip"

checkpoint = load_from_hub(repo_id, filename)

# 2. Load the trained model
model = PPO.load(checkpoint)

# 3. Create the environment and watch the agent play!
env = gym.make("LunarLander-v2", render_mode="human")
obs, info = env.reset()

for i in range(1000):
    action, _states = model.predict(obs, deterministic=True)
    obs, reward, terminated, truncated, info = env.step(action)
    
    if terminated or truncated:
        obs, info = env.reset()

env.close()