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: 213.83 +/- 3.18
            name: mean_reward
            verified: false

PPO Agent Playing LunarLander-v2

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

Hyperparameters

hp_seed: 5678
hp_total_timesteps: 500000
hp_learning_rate: 0.001
hp_min_learning_rate: 1.5e-06
hp_num_envs: 12
hp_num_steps: 256
hp_anneal_lr: True
hp_gae: True
hp_gamma: 0.99
hp_gae_lambda: 0.95
hp_minibatch_size: 256
hp_update_epochs: 4
hp_norm_adv: True
hp_clip_coef: 0.2
hp_clip_vloss: True
hp_ent_coef: 0.01
hp_vf_coef: 0.5
hp_max_grad_norm: 0.5
hp_target_kl: None