Views
No views yet
cleanrl package with the following command:pip install "cleanrl[jax,envpool,atari]"
python -m cleanrl_utils.enjoy --exp-name sebulba_ppo_envpool --env-id WizardOfWor-v51curl -OL https://huggingface.co/cleanrl/WizardOfWor-v5-sebulba_ppo_envpool-seed1/raw/main/sebulba_ppo_envpool.py
2curl -OL https://huggingface.co/cleanrl/WizardOfWor-v5-sebulba_ppo_envpool-seed1/raw/main/pyproject.toml
3curl -OL https://huggingface.co/cleanrl/WizardOfWor-v5-sebulba_ppo_envpool-seed1/raw/main/poetry.lock
4poetry install --all-extras
5python sebulba_ppo_envpool.py --actor-device-ids 0 --learner-device-ids 1 2 3 4 --params-queue-timeout 0.02 --track --save-model --upload-model --hf-entity cleanrl --env-id WizardOfWor-v5 --seed 11{'actor_device_ids': [0],
2 'anneal_lr': True,
3 'async_batch_size': 16,
4 'async_update': 4,
5 'batch_size': 8192,
6 'capture_video': False,
7 'clip_coef': 0.1,
8 'cuda': True,
9 'ent_coef': 0.01,
10 'env_id': 'WizardOfWor-v5',
11 'exp_name': 'sebulba_ppo_envpool',
12 'gae_lambda': 0.95,
13 'gamma': 0.99,
14 'hf_entity': 'cleanrl',
15 'learner_device_ids': [1, 2, 3, 4],
16 'learning_rate': 0.00025,
17 'max_grad_norm': 0.5,
18 'minibatch_size': 2048,
19 'norm_adv': True,
20 'num_actor_threads': 1,
21 'num_envs': 64,
22 'num_minibatches': 4,
23 'num_steps': 128,
24 'num_updates': 6103,
25 'params_queue_timeout': 0.02,
26 'profile': False,
27 'save_model': True,
28 'seed': 1,
29 'target_kl': None,
30 'test_actor_learner_throughput': False,
31 'torch_deterministic': True,
32 'total_timesteps': 50000000,
33 'track': True,
34 'update_epochs': 4,
35 'upload_model': True,
36 'vf_coef': 0.5,
37 'wandb_entity': None,
38 'wandb_project_name': 'cleanRL'}