Quentin Gallouédec
Initial commit
bce0eed
metadata
library_name: stable-baselines3
tags:
  - ReacherBulletEnv-v0
  - deep-reinforcement-learning
  - reinforcement-learning
  - stable-baselines3
model-index:
  - name: A2C
    results:
      - task:
          type: reinforcement-learning
          name: reinforcement-learning
        dataset:
          name: ReacherBulletEnv-v0
          type: ReacherBulletEnv-v0
        metrics:
          - type: mean_reward
            value: 3.06 +/- 11.48
            name: mean_reward
            verified: false

A2C Agent playing ReacherBulletEnv-v0

This is a trained model of a A2C agent playing ReacherBulletEnv-v0 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 a2c --env ReacherBulletEnv-v0 -orga qgallouedec -f logs/
python -m rl_zoo3.enjoy --algo a2c --env ReacherBulletEnv-v0  -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 a2c --env ReacherBulletEnv-v0 -orga qgallouedec -f logs/
python -m rl_zoo3.enjoy --algo a2c --env ReacherBulletEnv-v0  -f logs/

Training (with the RL Zoo)

python -m rl_zoo3.train --algo a2c --env ReacherBulletEnv-v0 -f logs/
# Upload the model and generate video (when possible)
python -m rl_zoo3.push_to_hub --algo a2c --env ReacherBulletEnv-v0 -f logs/ -orga qgallouedec

Hyperparameters

OrderedDict([('ent_coef', 0.0),
             ('gae_lambda', 0.9),
             ('gamma', 0.99),
             ('learning_rate', 'lin_0.0008'),
             ('max_grad_norm', 0.5),
             ('n_envs', 4),
             ('n_steps', 8),
             ('n_timesteps', 2000000.0),
             ('normalize', True),
             ('normalize_advantage', False),
             ('policy', 'MlpPolicy'),
             ('policy_kwargs', 'dict(log_std_init=-2, ortho_init=False)'),
             ('use_rms_prop', True),
             ('use_sde', True),
             ('vf_coef', 0.4),
             ('normalize_kwargs', {'norm_obs': True, 'norm_reward': False})])