tqc-PandaStack-v1 / README.md
qgallouedec's picture
qgallouedec HF staff
Add link to Panda Gym paper (#1)
91037b5
metadata
library_name: stable-baselines3
tags:
  - PandaStack-v1
  - deep-reinforcement-learning
  - reinforcement-learning
  - stable-baselines3
model-index:
  - name: TQC
    results:
      - metrics:
          - type: mean_reward
            value: '-99.90 +/- 0.30'
            name: mean_reward
        task:
          type: reinforcement-learning
          name: reinforcement-learning
        dataset:
          name: PandaStack-v1
          type: PandaStack-v1

TQC Agent playing PandaStack-v1

This is a trained model of a TQC agent playing PandaStack-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

# Download model and save it into the logs/ folder
python -m rl_zoo3.load_from_hub --algo tqc --env PandaStack-v1 -orga sb3 -f logs/
python enjoy.py --algo tqc --env PandaStack-v1  -f logs/

Training (with the RL Zoo)

python train.py --algo tqc --env PandaStack-v1 -f logs/
# Upload the model and generate video (when possible)
python -m rl_zoo3.push_to_hub --algo tqc --env PandaStack-v1 -f logs/ -orga sb3

Hyperparameters

OrderedDict([('batch_size', 2048),
             ('buffer_size', 1000000),
             ('env_wrapper', 'sb3_contrib.common.wrappers.TimeFeatureWrapper'),
             ('gamma', 0.95),
             ('learning_rate', 0.001),
             ('n_timesteps', 1000000.0),
             ('policy', 'MultiInputPolicy'),
             ('policy_kwargs', 'dict(net_arch=[512, 512, 512], n_critics=2)'),
             ('replay_buffer_class', 'HerReplayBuffer'),
             ('replay_buffer_kwargs',
              "dict( online_sampling=True, goal_selection_strategy='future', "
              'n_sampled_goal=4, )'),
             ('tau', 0.05),
             ('normalize', False)])

Panda Gym environments: arxiv.org/abs/2106.13687