Update README.md
Browse files
README.md
CHANGED
|
@@ -26,4 +26,29 @@ model-index:
|
|
| 26 |
# **RND with CNN** Agent playing **SuperMarioBros-v0**
|
| 27 |
This is a trained model of a **RND-CNN** agent playing **SuperMarioBros-v0** .
|
| 28 |
To learn to use this model and train yours check this notebook on kaggle: https://www.kaggle.com/code/syedjarullahhisham/drl-extra-personal-unit-5-rnd-super-mario-bros
|
| 29 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 26 |
# **RND with CNN** Agent playing **SuperMarioBros-v0**
|
| 27 |
This is a trained model of a **RND-CNN** agent playing **SuperMarioBros-v0** .
|
| 28 |
To learn to use this model and train yours check this notebook on kaggle: https://www.kaggle.com/code/syedjarullahhisham/drl-extra-personal-unit-5-rnd-super-mario-bros
|
| 29 |
+
|
| 30 |
+
# HyperParameters:
|
| 31 |
+
|
| 32 |
+
```
|
| 33 |
+
"trainmethod": "RND",
|
| 34 |
+
"envid": "SuperMarioBros-v0",
|
| 35 |
+
"maxstepperepisode": 18000,
|
| 36 |
+
"learningrate": 0.0001,
|
| 37 |
+
"numenv": 128,
|
| 38 |
+
"numstep": 128,
|
| 39 |
+
"gamma": 0.999,
|
| 40 |
+
"intgamma": 0.99,
|
| 41 |
+
"lambda": 0.95,
|
| 42 |
+
"usegae": true,
|
| 43 |
+
"clipgradnorm": 0.5,
|
| 44 |
+
"entropy": 0.001,
|
| 45 |
+
"epoch": 4,
|
| 46 |
+
"minibatch": 4,
|
| 47 |
+
"ppoeps": 0.1,
|
| 48 |
+
"extcoef": 5.0,
|
| 49 |
+
"intcoef": 1.0,
|
| 50 |
+
"stickyaction": true,
|
| 51 |
+
"actionprob": 0.25,
|
| 52 |
+
"lifedone": false,
|
| 53 |
+
"obsnormstep": 50
|
| 54 |
+
```
|