Update README.md
Browse files
README.md
CHANGED
|
@@ -21,7 +21,11 @@ license: afl-3.0
|
|
| 21 |
---
|
| 22 |
|
| 23 |
# **DQN** Agent playing **MountainCar-v0**
|
| 24 |
-
This is a trained model of a **DQN** agent playing **MountainCar-v0
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
|
| 26 |
Parameters:
|
| 27 |
```python
|
|
|
|
| 21 |
---
|
| 22 |
|
| 23 |
# **DQN** Agent playing **MountainCar-v0**
|
| 24 |
+
This is a trained model of a **DQN** agent playing **MountainCar-v0**.
|
| 25 |
+
We train a three-layer MLP as the Q-network.
|
| 26 |
+
We store the transitions in a replay buffer.
|
| 27 |
+
After the network converges, we stop training and validate its performance in
|
| 28 |
+
comparison to a random baseline.
|
| 29 |
|
| 30 |
Parameters:
|
| 31 |
```python
|