| --- |
| tags: |
| - Pixelcopter-PLE-v0 |
| - reinforce |
| - reinforcement-learning |
| - custom-implementation |
| - deep-rl-class |
| model-index: |
| - name: Reinforce-PixelCopter-v2 |
| results: |
| - task: |
| type: reinforcement-learning |
| name: reinforcement-learning |
| dataset: |
| name: Pixelcopter-PLE-v0 |
| type: Pixelcopter-PLE-v0 |
| metrics: |
| - type: mean_reward |
| value: 18.50 +/- 21.02 |
| name: mean_reward |
| verified: false |
| --- |
| <br> |
| pixelcopter_hyperparameters = { <br> |
| "h_size": 64, <br> |
| "n_training_episodes": 50000, <br> |
| "n_evaluation_episodes": 10, <br> |
| "max_t": 10000, <br> |
| "gamma": 0.99, <br> |
| "lr": 1e-4, <br> |
| "env_id": env_id, <br> |
| "state_space": s_size, <br> |
| "action_space": a_size, <br> |
| }<br> |
| |
