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)
TODO: Add your code
!apt install swig cmake
virtual screen to be able to render the environment
!sudo apt-get update
!sudo apt-get install -y python3-opengl
!apt install ffmpeg
!apt install xvfb
!pip3 install pyvirtualdisplay
import os
os.kill(os.getpid(), 9)
Virtual display
from pyvirtualdisplay import Display
virtual_display = Display(visible=0, size=(1400, 900))
virtual_display.start()
!pip install gymnasium[box2d]
!pip install stable-baselines3[extra]
!pip install huggingface_sb3
from huggingface_sb3 import load_from_hub, package_to_hub
from huggingface_hub import notebook_loginub
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
import gymnasium as gym
First, we create our environment called LunarLander-v2
environ = gym.make("LunarLander-v2")
Then we reset this environment
observation, info = environ.reset()
for i in range(20):
Take a random action
action = environ.action_space.sample()
print("Action taken:", action)
Do this action in the environment and get
next_state, reward, terminated, truncated and info
observation, reward, terminated, truncated, info = environ.step(action)
If the game is terminated (in our case we land, crashed) or truncated (timeout)
if terminated or truncated:
# Reset the environment
print("Environment is reset")
observation, info = environ.reset()
environ.close()
environ = gym.make("LunarLander-v2")
environ.reset()
print("OBSERVATION SPACE \n")
print("Observation Space Shape", environ.observation_space.shape)
Get a random observation
print("Sample observation", environ.observation_space.sample())
print("\n ACTION SPACE \n")
print("Action Space Shape", environ.action_space.n)
print("Action Space Sample", environ.action_space.sample())
Create the environment
environ = make_vec_env('LunarLander-v2', n_envs=16)
Create environment
environ = gym.make('LunarLander-v2')
Instantiate the agent
model = PPO(
policy='MlpPolicy',
env=environ,
n_steps=1024,
batch_size=64,
n_epochs=4,
gamma=0.999,
gae_lambda=0.98,
ent_coef=0.01,
verbose=1
)
Train the agent
model.learn(total_timesteps=1000000)
Save the model
model_name = "ppo-LunarLander-v2"
model.save(model_name)
Create a new environment for evaluation
eval_env = Monitor(gym.make("LunarLander-v2", render_mode='rgb_array'))
Evaluate the model with 10 evaluation episodes and deterministic=True
mean_reward, std_reward = evaluate_policy(model, eval_env, n_eval_episodes=10, deterministic=True)
Print the results
print(f"mean_reward={mean_reward:.2f} +/- {std_reward}")
import gymnasium as gym
from stable_baselines3 import PPO
from stable_baselines3.common.vec_env import DummyVecEnv
from stable_baselines3.common.env_util import make_vec_env
from huggingface_sb3 import package_to_hub
PLACE the variables you've just defined two cells above
Define the name of the environment
env_id = "LunarLander-v2"
TODO: Define the model architecture we used
model_architecture = "PPO"
Define a repo_id
repo_id is the id of the model repository from the Hugging Face Hub (repo_id = {organization}/{repo_name} for instance ThomasSimonini/ppo-LunarLander-v2
CHANGE WITH YOUR REPO ID
repo_id = "dns08/LunarLander-v2" # Change with your repo id, you can't push with mine 😄
Define the commit message
commit_message = "Upload PPO LunarLander-v2 trained agent"
Create the evaluation env and set the render_mode="rgb_array"
eval_env = DummyVecEnv([lambda: gym.make(env_id, render_mode="rgb_array")])
PLACE the package_to_hub function you've just filled here
package_to_hub(model=model, # Our trained model
model_name=model_name, # The name of our trained model
model_architecture=model_architecture, # The model architecture we used: in our case PPO
env_id=env_id, # Name of the environment
eval_env=eval_env, # Evaluation Environment
repo_id=repo_id, # id of the model repository from the Hugging Face Hub (repo_id = {organization}/{repo_name} for instance ThomasSimonini/ppo-LunarLander-v2
commit_message=commit_message)
!pip install huggingface_sb3
!pip install gymnasium stable-baselines3 huggingface_sb3
from stable_baselines3 import PPO
from huggingface_sb3 import load_from_hub
repo_id = "dns08/LunarLander-v2" # The repo_id
filename = "ppo-LunarLander-v2.zip" # The model filename.zip
When the model was trained on Python 3.8 the pickle protocol is 5
But Python 3.6, 3.7 use protocol 4
In order to get compatibility we need to:
1. Install pickle5 (we done it at the beginning of the colab)
2. Create a custom empty object we pass as parameter to PPO.load()
custom_objects = {
"learning_rate": 0.0,
"lr_schedule": lambda _: 0.0,
"clip_range": lambda _: 0.0,
}
checkpoint = load_from_hub(repo_id, filename)
model = PPO.load(checkpoint, custom_objects=custom_objects, print_system_info=True)
!apt-get install swig
!pip install box2d-py
!pip install gymnasium[box2d]
from stable_baselines3.common.monitor import Monitor
from stable_baselines3.common.evaluation import evaluate_policy
import gymnasium as gym
Create and monitor the evaluation environment
eval_env = Monitor(gym.make("LunarLander-v2"))
Evaluate the trained model
mean_reward, std_reward = evaluate_policy(model, eval_env, n_eval_episodes=10, deterministic=True)
Print the results
print(f"mean_reward={mean_reward:.2f} +/- {std_reward}")
