Edit model card

(CleanRL) PPO Agent Playing Pitfall-v5

This is a trained model of a PPO agent playing Pitfall-v5. 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[ppo_atari_envpool_async_jax_scan_impalanet_machado]"
python -m cleanrl_utils.enjoy --exp-name ppo_atari_envpool_async_jax_scan_impalanet_machado --env-id Pitfall-v5

Please refer to the documentation for more detail.

Command to reproduce the training

curl -OL https://huggingface.co/cleanrl/Pitfall-v5-ppo_atari_envpool_async_jax_scan_impalanet_machado-seed1/raw/main/ppo_atari_envpool_async_jax_scan_impalanet_machado.py
curl -OL https://huggingface.co/cleanrl/Pitfall-v5-ppo_atari_envpool_async_jax_scan_impalanet_machado-seed1/raw/main/pyproject.toml
curl -OL https://huggingface.co/cleanrl/Pitfall-v5-ppo_atari_envpool_async_jax_scan_impalanet_machado-seed1/raw/main/poetry.lock
poetry install --all-extras
python ppo_atari_envpool_async_jax_scan_impalanet_machado.py --track --wandb-project-name envpool-atari --save-model --upload-model --hf-entity cleanrl --env-id Pitfall-v5 --seed 1

Hyperparameters

{'anneal_lr': True,
 'async_batch_size': 16,
 'batch_size': 2048,
 'capture_video': False,
 'clip_coef': 0.1,
 'cuda': True,
 'ent_coef': 0.01,
 'env_id': 'Pitfall-v5',
 'exp_name': 'ppo_atari_envpool_async_jax_scan_impalanet_machado',
 'gae': True,
 'gae_lambda': 0.95,
 'gamma': 0.99,
 'hf_entity': 'cleanrl',
 'learning_rate': 0.00025,
 'max_grad_norm': 0.5,
 'minibatch_size': 1024,
 'norm_adv': True,
 'num_envs': 64,
 'num_minibatches': 2,
 'num_steps': 32,
 'num_updates': 24414,
 'save_model': True,
 'seed': 1,
 'target_kl': None,
 'torch_deterministic': True,
 'total_timesteps': 50000000,
 'track': True,
 'update_epochs': 2,
 'upload_model': True,
 'vf_coef': 0.5,
 'wandb_entity': None,
 'wandb_project_name': 'envpool-atari'}
Downloads last month

-

Downloads are not tracked for this model. How to track
Video Preview
loading

Evaluation results