README.md

metadata

tags:
  - PongNoFrameskip-v4
  - deep-reinforcement-learning
  - reinforcement-learning
  - custom-implementation
library_name: cleanrl
model-index:
  - name: DQPN_freq
    results:
      - task:
          type: reinforcement-learning
          name: reinforcement-learning
        dataset:
          name: PongNoFrameskip-v4
          type: PongNoFrameskip-v4
        metrics:
          - type: mean_reward
            value: 19.03 +/- 0.00
            name: mean_reward
            verified: false

(CleanRL) DQPN_freq Agent Playing PongNoFrameskip-v4

This is a trained model of a DQPN_freq agent playing PongNoFrameskip-v4. The model was trained by using CleanRL and the most up-to-date training code can be found here.

Get Started

To use this model, please install the cleanrl package with the following command:

pip install "cleanrl[DQPN_x1]"
python -m cleanrl_utils.enjoy --exp-name DQPN_x1 --env-id PongNoFrameskip-v4

Please refer to the documentation for more detail.

Command to reproduce the training

curl -OL https://huggingface.co/pfunk/PongNoFrameskip-v4-DQPN_x1-seed3/raw/main/dqpn_freq_atari.py
curl -OL https://huggingface.co/pfunk/PongNoFrameskip-v4-DQPN_x1-seed3/raw/main/pyproject.toml
curl -OL https://huggingface.co/pfunk/PongNoFrameskip-v4-DQPN_x1-seed3/raw/main/poetry.lock
poetry install --all-extras
python dqpn_freq_atari.py --track --wandb-entity pfunk --wandb-project-name dqpn --capture-video true --save-model true --upload-model true --hf-entity pfunk --exp-name DQPN_x1 --target-network-frequency 1000 --policy-network-frequency 1000 --seed 3

Hyperparameters

{'alg_type': 'dqpn_freq_atari.py',
 'batch_size': 32,
 'buffer_size': 1000000,
 'capture_video': True,
 'cuda': True,
 'double_learning': False,
 'end_e': 0.05,
 'env_id': 'PongNoFrameskip-v4',
 'exp_name': 'DQPN_x1',
 'exploration_fraction': 0.2,
 'gamma': 0.99,
 'hf_entity': 'pfunk',
 'learning_rate': 0.0001,
 'learning_starts': 10000,
 'max_gradient_norm': inf,
 'policy_network_frequency': 1000,
 'policy_tau': 1.0,
 'save_model': True,
 'seed': 3,
 'start_e': 1.0,
 'target_network_frequency': 1000,
 'target_tau': 1.0,
 'torch_deterministic': True,
 'total_timesteps': 10000000,
 'track': True,
 'train_frequency': 1,
 'upload_model': True,
 'wandb_entity': 'pfunk',
 'wandb_project_name': 'dqpn'}