You can deploy this model and start earning money today!

HumanCompatibleAI/
ppo-seals-Swimmer-v1

Views

No views yet

PPO Agent playing seals/Swimmer-v1

This is a trained model of a PPO agent playing seals/Swimmer-v1 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

Install the RL Zoo (with SB3 and SB3-Contrib):

pip install rl_zoo3

# Download model and save it into the logs/ folder
python -m rl_zoo3.load_from_hub --algo ppo --env seals/Swimmer-v1 -orga HumanCompatibleAI -f logs/
python -m rl_zoo3.enjoy --algo ppo --env seals/Swimmer-v1  -f logs/

If you installed the RL Zoo3 via pip (pip install rl_zoo3), from anywhere you can do:

python -m rl_zoo3.load_from_hub --algo ppo --env seals/Swimmer-v1 -orga HumanCompatibleAI -f logs/
python -m rl_zoo3.enjoy --algo ppo --env seals/Swimmer-v1  -f logs/

Training (with the RL Zoo)

python -m rl_zoo3.train --algo ppo --env seals/Swimmer-v1 -f logs/
# Upload the model and generate video (when possible)
python -m rl_zoo3.push_to_hub --algo ppo --env seals/Swimmer-v1 -f logs/ -orga HumanCompatibleAI

Hyperparameters

python

OrderedDict([('batch_size', 8),
             ('clip_range', 0.1),
             ('ent_coef', 5.167107294612664e-08),
             ('gae_lambda', 0.95),
             ('gamma', 0.999),
             ('learning_rate', 0.0001214437022727675),
             ('max_grad_norm', 2),
             ('n_epochs', 20),
             ('n_steps', 2048),
             ('n_timesteps', 1000000.0),
             ('normalize',
              {'gamma': 0.999, 'norm_obs': False, 'norm_reward': True}),
             ('policy', 'MlpPolicy'),
             ('policy_kwargs',
              {'activation_fn': <class 'torch.nn.modules.activation.Tanh'>,
               'features_extractor_class': <class 'imitation.policies.base.NormalizeFeaturesExtractor'>,
               'net_arch': [{'pi': [64, 64], 'vf': [64, 64]}]}),
             ('vf_coef', 0.6162112311062333),
             ('normalize_kwargs',
              {'norm_obs': {'gamma': 0.999,
                            'norm_obs': False,
                            'norm_reward': True},
               'norm_reward': False})])

Environment Arguments

{'render_mode': 'rgb_array'}

HumanCompatibleAI/ppo-seals-Swimmer-v1