Update README.md
Browse files
README.md
CHANGED
|
@@ -22,40 +22,40 @@ model-index:
|
|
| 22 |
verified: false
|
| 23 |
---
|
| 24 |
|
| 25 |
-
|
| 26 |
|
| 27 |
-
|
| 28 |
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
|
| 34 |
-
|
| 35 |
-
|
| 36 |
-
|
| 37 |
-
|
| 38 |
-
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
|
| 46 |
-
|
| 47 |
-
|
| 48 |
-
|
| 49 |
-
|
| 50 |
-
|
| 51 |
-
|
| 52 |
-
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
| 61 |
|
|
|
|
| 22 |
verified: false
|
| 23 |
---
|
| 24 |
|
| 25 |
+
# PPO Agent Playing CartPole-v1
|
| 26 |
|
| 27 |
+
This is a trained model of a PPO agent playing CartPole-v1.
|
| 28 |
|
| 29 |
+
# Hyperparameters
|
| 30 |
+
```python
|
| 31 |
+
{'exp_name': 'unit8_part1',
|
| 32 |
+
'gym_id': 'CartPole-v1',
|
| 33 |
+
'learning_rate': 0.00025,
|
| 34 |
+
'seed': 1,
|
| 35 |
+
'total_timesteps': 50000,
|
| 36 |
+
'torch_deterministic': True,
|
| 37 |
+
'cuda': True,
|
| 38 |
+
'track': False,
|
| 39 |
+
'wandb_project_name': 'ppo-implementation-details',
|
| 40 |
+
'wandb_entity': None,
|
| 41 |
+
'capture_video': False,
|
| 42 |
+
'num_envs': 4,
|
| 43 |
+
'num_steps': 128,
|
| 44 |
+
'anneal_lr': True,
|
| 45 |
+
'gae': True,
|
| 46 |
+
'gamma': 0.99,
|
| 47 |
+
'gae_lambda': 0.95,
|
| 48 |
+
'num_minibatches': 4,
|
| 49 |
+
'update_epochs': 4,
|
| 50 |
+
'norm_adv': True,
|
| 51 |
+
'clip_coef': 0.2,
|
| 52 |
+
'clip_vloss': True,
|
| 53 |
+
'ent_coef': 0.01,
|
| 54 |
+
'vf_coef': 0.5,
|
| 55 |
+
'max_grad_norm': 0.5,
|
| 56 |
+
'target_kl': None,
|
| 57 |
+
'repo_id': 'tmoroder/ppo-CartPole-v1',
|
| 58 |
+
'batch_size': 512,
|
| 59 |
+
'minibatch_size': 128}
|
| 60 |
+
```
|
| 61 |
|