README.md

metadata

library_name: stable-baselines3
tags:
  - LunarLander-v2
  - deep-reinforcement-learning
  - reinforcement-learning
  - stable-baselines3
model-index:
  - name: PPO
    results:
      - task:
          type: reinforcement-learning
          name: reinforcement-learning
        dataset:
          name: LunarLander-v2
          type: LunarLander-v2
        metrics:
          - type: mean_reward
            value: 242.08 +/- 19.81
            name: mean_reward
            verified: false

PPO Agent playing LunarLander-v2

This is a trained model of a PPO agent playing LunarLander-v2 using the stable-baselines3 library.

Usage (with Stable-baselines3)

import gymnasium as gym
from time import sleep
from huggingface_sb3 import package_to_hub
from stable_baselines3 import PPO
from stable_baselines3.common.env_util import make_vec_env
from stable_baselines3.common.evaluation import evaluate_policy
from stable_baselines3.common.monitor import Monitor
from stable_baselines3.common.vec_env import DummyVecEnv

# Create the environment
env = make_vec_env("LunarLander-v2", n_envs=16)

# We added some parameters to accelerate the training
model = PPO(
    policy="MlpPolicy",
    env=env,
    n_steps=1024,
    batch_size=64,
    n_epochs=4,
    gamma=0.999,
    gae_lambda=0.98,
    ent_coef=0.01,
    verbose=1,
)

# Train it for 1,000,000 timesteps
model.learn(total_timesteps=1000000)
# Save the model
model.save(model_name)

# Test the model
# model = PPO.load(model_name)
eval_env = Monitor(gym.make("LunarLander-v2"))
mean_reward, std_reward = evaluate_policy(model, eval_env, n_eval_episodes=10, deterministic=True)
print(f"mean_reward={mean_reward:.2f} +/- {std_reward}")

# Visualize the model
env = gym.make("LunarLander-v2", render_mode='human')

state, _ = env.reset()
stop = False

while not stop:
    action, _ = model.predict(state)
    state, reward, terminated, truncated, info = env.step(action)
    stop = terminated or truncated
    env.render()
    sleep(0.05)

    if terminated or truncated:
        observation, info = env.reset()

env.close()
...