File size: 851 Bytes
d4a51fa
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
fb744f0
d4a51fa
 
fb744f0
d4a51fa
 
 
fc7a661
 
 
 
 
d4a51fa
fb744f0
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
---
tags:
- MountainCar-v0
- deep-reinforcement-learning
- reinforcement-learning
model-index:
- name: DQN
  results:
  - task:
      type: reinforcement-learning
      name: reinforcement-learning
    dataset:
      name: MountainCar-v0
      type: MountainCar-v0
    metrics:
    - type: mean_reward
      value: '-120.10 +/- 19.30'
      name: mean_reward
      verified: false
license: afl-3.0
---

# **DQN** Agent playing **MountainCar-v0**
This is a trained model of a **DQN** agent playing **MountainCar-v0**. 
We train a three-layer MLP as the Q-network. 
We store the transitions in a replay buffer.
After the network converges, we stop training and validate its performance in
comparison to a random baseline.

Parameters:
```python
hidden_size = 64
gamma = 0.99
epsilon_decay = 0.999
buffer_size = 10000
batch_size = 64
episodes = 10000
```