araffin's picture
Upload README.md with huggingface_hub
b5a2145
metadata
library_name: stable-baselines3
tags:
  - BipedalWalkerHardcore-v3
  - deep-reinforcement-learning
  - reinforcement-learning
  - stable-baselines3
model-index:
  - name: PPO
    results:
      - metrics:
          - type: mean_reward
            value: 48.61 +/- 120.68
            name: mean_reward
        task:
          type: reinforcement-learning
          name: reinforcement-learning
        dataset:
          name: BipedalWalkerHardcore-v3
          type: BipedalWalkerHardcore-v3

PPO Agent playing BipedalWalkerHardcore-v3

This is a trained model of a PPO agent playing BipedalWalkerHardcore-v3 using the stable-baselines3 library and the RL Zoo.

The RL Zoo is a training framework for Stable Baselines3 reinforcement learning agents, with hyperparameter optimization and pre-trained agents included.

Usage (with SB3 RL Zoo)

RL Zoo: https://github.com/DLR-RM/rl-baselines3-zoo
SB3: https://github.com/DLR-RM/stable-baselines3
SB3 Contrib: https://github.com/Stable-Baselines-Team/stable-baselines3-contrib

# Download model and save it into the logs/ folder
python -m rl_zoo3.load_from_hub --algo ppo --env BipedalWalkerHardcore-v3 -orga sb3 -f logs/
python enjoy.py --algo ppo --env BipedalWalkerHardcore-v3  -f logs/

Training (with the RL Zoo)

python train.py --algo ppo --env BipedalWalkerHardcore-v3 -f logs/
# Upload the model and generate video (when possible)
python -m rl_zoo3.push_to_hub --algo ppo --env BipedalWalkerHardcore-v3 -f logs/ -orga sb3

Hyperparameters

OrderedDict([('batch_size', 64),
             ('clip_range', 'lin_0.2'),
             ('ent_coef', 0.001),
             ('gae_lambda', 0.95),
             ('gamma', 0.99),
             ('learning_rate', 'lin_2.5e-4'),
             ('n_envs', 16),
             ('n_epochs', 10),
             ('n_steps', 2048),
             ('n_timesteps', 100000000.0),
             ('normalize', True),
             ('policy', 'MlpPolicy'),
             ('normalize_kwargs', {'norm_obs': True, 'norm_reward': False})])