README.md

metadata

tags:
  - LunarLander-v2
  - ppo
  - deep-reinforcement-learning
  - reinforcement-learning
  - custom-implementation
  - deep-rl-course
model-index:
  - name: PPO
    results:
      - task:
          type: reinforcement-learning
          name: reinforcement-learning
        dataset:
          name: LunarLander-v2
          type: LunarLander-v2
        metrics:
          - type: mean_reward
            value: 0.60 +/- 1.00
            name: mean_reward
            verified: false

PPO Agent Playing LunarLander-v2

This is a trained model of a PPO agent playing LunarLander-v2.

Hyperparameters

{'exp_name': '3million'
'seed': 1
'torch_deterministic': True
'cuda': True
'track': True
'wandb_project_name': 'LunarLanderProject'
'wandb_entity': None
'capture_video': True
'env_id': 'LunarLander-v2'
'total_timesteps': 3000000
'learning_rate': 0.0004
'num_envs': 4
'num_steps': 128
'anneal_lr': False
'gae': True
'gamma': 0.999
'gae_lambda': 0.98
'num_minibatches': 4
'update_epochs': 16
'norm_adv': True
'clip_coef': 0.2
'clip_vloss': True
'ent_coef': 0.01
'vf_coef': 0.5
'max_grad_norm': 0.5
'target_kl': None
'num_hidden_layers': 2
'upload_to_hf': True
'repo_id': 'candrews1971/ppo-LunarLander-v2.1'
'batch_size': 512
'minibatch_size': 128}