Hugging Face's logo

cleanrl
/
WizardOfWor-v5-cleanba_impala_envpool_impala_atari_wrapper_a0_l1_d4-seed3

Reinforcement Learning
TensorBoard
cleanrl
WizardOfWor-v5
deep-reinforcement-learning
custom-implementation
Eval Results
Model card Files Metrics Community
WizardOfWor-v5-cleanba_impala_envpool_impala_atari_wrapper_a0_l1_d4-seed3 / README.md
vwxyzjn's picture
vwxyzjn
pushing model
af6e0b4
preview code
|
raw
history blame contribute delete
3.06 kB
metadata
tags:
  - WizardOfWor-v5
  - deep-reinforcement-learning
  - reinforcement-learning
  - custom-implementation
library_name: cleanrl
model-index:
  - name: PPO
    results:
      - task:
          type: reinforcement-learning
          name: reinforcement-learning
        dataset:
          name: WizardOfWor-v5
          type: WizardOfWor-v5
        metrics:
          - type: mean_reward
            value: 8210.00 +/- 6066.04
            name: mean_reward
            verified: false

(CleanRL) PPO Agent Playing WizardOfWor-v5

This is a trained model of a PPO agent playing WizardOfWor-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[jax,envpool,atari]"
python -m cleanrl_utils.enjoy --exp-name cleanba_impala_envpool_impala_atari_wrapper_a0_l1_d4 --env-id WizardOfWor-v5

Please refer to the documentation for more detail.

Command to reproduce the training 

curl -OL https://huggingface.co/cleanrl/WizardOfWor-v5-cleanba_impala_envpool_impala_atari_wrapper_a0_l1_d4-seed3/raw/main/cleanba_impala_envpool_impala_atari_wrapper.py
curl -OL https://huggingface.co/cleanrl/WizardOfWor-v5-cleanba_impala_envpool_impala_atari_wrapper_a0_l1_d4-seed3/raw/main/pyproject.toml
curl -OL https://huggingface.co/cleanrl/WizardOfWor-v5-cleanba_impala_envpool_impala_atari_wrapper_a0_l1_d4-seed3/raw/main/poetry.lock
poetry install --all-extras
python cleanba_impala_envpool_impala_atari_wrapper.py --exp-name cleanba_impala_envpool_impala_atari_wrapper_a0_l1_d4 --distributed --learner-device-ids 1 --local-num-envs 30 --track --wandb-project-name cleanba --save-model --upload-model --hf-entity cleanrl --env-id WizardOfWor-v5 --seed 3

Hyperparameters 

{'actor_device_ids': [0],
 'actor_devices': ['gpu:0'],
 'anneal_lr': True,
 'async_batch_size': 30,
 'async_update': 1,
 'batch_size': 2400,
 'capture_video': False,
 'cuda': True,
 'distributed': True,
 'ent_coef': 0.01,
 'env_id': 'WizardOfWor-v5',
 'exp_name': 'cleanba_impala_envpool_impala_atari_wrapper_a0_l1_d4',
 'gamma': 0.99,
 'global_learner_decices': ['gpu:1', 'gpu:3', 'gpu:5', 'gpu:7'],
 'hf_entity': 'cleanrl',
 'learner_device_ids': [1],
 'learner_devices': ['gpu:1'],
 'learning_rate': 0.00025,
 'local_batch_size': 600,
 'local_minibatch_size': 300,
 'local_num_envs': 30,
 'local_rank': 0,
 'max_grad_norm': 0.5,
 'minibatch_size': 1200,
 'num_envs': 120,
 'num_minibatches': 2,
 'num_steps': 20,
 'num_updates': 20833,
 'profile': False,
 'save_model': True,
 'seed': 3,
 'target_kl': None,
 'test_actor_learner_throughput': False,
 'torch_deterministic': True,
 'total_timesteps': 50000000,
 'track': True,
 'upload_model': True,
 'vf_coef': 0.5,
 'wandb_entity': None,
 'wandb_project_name': 'cleanba',
 'world_size': 4}