|
|
--- |
|
|
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: 230.81 +/- 20.92 |
|
|
name: mean_reward |
|
|
verified: false |
|
|
--- |
|
|
|
|
|
# PPO Agent Playing LunarLander-v2 |
|
|
|
|
|
This is a trained model of a PPO agent playing LunarLander-v2. |
|
|
|
|
|
# Hyperparameters |
|
|
```python |
|
|
{'path': '/content/drive/MyDrive/Colab Notebooks/HuggingFace/RL/Unit08' |
|
|
'name': 'ppo-LunaLander_1.pt' |
|
|
'env-id': 'LunarLander-v2' |
|
|
'agent_properties': {'num_layers': 2 |
|
|
'hidden': 128 |
|
|
'activation': 'Tanh'} |
|
|
'seed': '' |
|
|
'device': 'cuda' |
|
|
'total_timesteps': 100000 |
|
|
'num_steps': 32768 |
|
|
'batch_size': 64 |
|
|
'update_epochs': 2 |
|
|
'learning_rate': 1e-05 |
|
|
'lr_schedule': 'Exp' |
|
|
'lr_final': 1e-06 |
|
|
'gamma': 0.995 |
|
|
'gae_lambda': 0.99 |
|
|
'norm_adv': 'True' |
|
|
'clip_coef': 0.2 |
|
|
'clip_vloss': 'False' |
|
|
'entropy_loss_coef': 0.01 |
|
|
'value_loss_coef': 0.5 |
|
|
'max_grad_norm': 0.5 |
|
|
'n_eval_episodes': 10} |
|
|
``` |
|
|
|
