buildthemachine/ppo-LunarLander-v2

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)

The PPO model is saved as "ppo-LunarLander-v2.zip".

There are two ways to load this model:

1. Directly load the model from huggingface. This requires the use of the 'load_from_hub' function

from huggingface_sb3 import load_from_hub
repo_id = "buildthemachine/ppo-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)

2. Directly loading the zip file:

model = PPO(policy = 'MlpPolicy',
            env = env,
            learning_rate = 3e-4,
            n_steps = 1024,
            batch_size = 64,
            n_epochs = 4,
            gamma = 0.999,
            gae_lambda = 0.98,
            ent_coef = 0.01,
            verbose=1)
model_name = "ppo-LunarLander-v2"
model = PPO.load((f"/content/drive/MyDrive/Colab Notebooks/RL_tutorial_model_save/{model_name}"))