Sahilsea/ppo-LunarLander-v3

PPO Agent playing LunarLander-v3

This is a trained model of a PPO agent playing LunarLander-v3 using the stable-baselines3 library.

Usage (with Stable-baselines3)

TODO: Add your code

import gymnasium as gym
from stable_baselines3 import PPO
from huggingface_sb3 import load_from_hub
from gymnasium.wrappers import RecordVideo
from IPython.display import Video, display
import glob
import os

# 1. Download and load your checkpoint
checkpoint = load_from_hub(
    repo_id="Sahilsea/ppo-LunarLander-v3",
    filename="ppo-LunarLander-v3.zip",
)
model = PPO.load(checkpoint)

# 2. Set up the environment with a video recorder wrapper
video_folder = "./videos"
env = gym.make("LunarLander-v3", render_mode="rgb_array")
env = RecordVideo(env, video_folder=video_folder, episode_trigger=lambda e: True)

# 3. Run one full landing episode
obs, info = env.reset()
done = False

while not done:
    action, _states = model.predict(obs, deterministic=True)
    obs, reward, terminated, truncated, info = env.step(action)
    done = terminated or truncated

env.close()

# 4. Display the recorded MP4 file
mp4_files = glob.glob(f"{video_folder}/*.mp4")
if mp4_files:
    display(Video(mp4_files[-1], embed=True, html_attributes="controls autoplay loop"))
else:
    print("No video file found.")
...