https://github.com/nderousseaux/lunarlander-v2-qlearning
My implementation of q-learning for lunar-landerv2
https://github.com/nderousseaux/lunarlander-v2-qlearning
experiment machine-learning tensorflow
Last synced: about 2 months ago
JSON representation
My implementation of q-learning for lunar-landerv2
- Host: GitHub
- URL: https://github.com/nderousseaux/lunarlander-v2-qlearning
- Owner: nderousseaux
- Created: 2023-06-18T20:18:53.000Z (about 3 years ago)
- Default Branch: main
- Last Pushed: 2023-06-27T15:53:12.000Z (about 3 years ago)
- Last Synced: 2025-02-12T18:34:13.533Z (over 1 year ago)
- Topics: experiment, machine-learning, tensorflow
- Language: Python
- Homepage:
- Size: 970 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Lunar Lander Q-Learning attempt
This is my attempt to solve the Lunar Lander problem using Q-Learning.
## How to use this code
All the parameters can be indicated on the command line like this:
```bash
python main.py --hidden_layers="[64, 64]" --learning_rate=0.001 --epoch=1000 --gamma=0.99 --epsilon=1.0 --epsilon_dec=0.999 --epsilon_end=0.01 --mem_size=100000 --batch_size=64 --activation_function="relu" --render_during_training=True --live_plot=True
```
If you not indicate any parameter, the default value will be used. All the defaults the parameters is in the top of the main.py file.
If there is already a model saved, (on the folder `models`), the program will skip the training and will play the game using the model.