milp_constrained_gan

KO | EN

gitlite — search

#hacktoberfest #ai #typescript #llm #rust #linux #docker #windows #awesome #cli #pytorch #ios #openai #open-source

milp_constrained_gan

(★ 9)

Using MILP to encode constraints of generated game levels.

milp_constrained_gan 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Video Game Level Repair via Mixed Integer Linear Programming

### Setup Instructions
- Install [CPLEX Optimization Studio and it Python API](https://developer.ibm.com/docloud/blog/2019/07/04/cplex-optimization-studio-for-students-and-academics/)
    - You need full version to solve large programs. (Researchers can use academic version).
- Unzip the file in a folder
- pip install -r requirements.txt
- Add the folder path to your PYTHONPATH

### Zelda domain
Training a GAN that learns the distribution of *Zelda* levels
- in the root folder of the project
- python launchers/zelda_gan_training.py --gan_experiment=\<path to save the samples and models\> --lvl_data=\<path to human authored levels>\
- If you have GPUs, you can use --cuda to enable GPUs.

Generate fixed levels
- in the root folder of the project
- python launchers/zelda_gan_generate.py --output_folder=\<path to save the generated levels\> --network_path=\<path to the save generator network\>
    - We have a pretrained model in default_samples
    
Visualize the generated levels
- in the root folder of the project
- python launchers/zelda_grid_visualize.py --lvl_path=\<path to the generated levels\> --output_folder=\<path to save the visualized levels\>

### PacMan domain
Training a GAN that learns the distribution of *PacMan* levels
- in the root folder of the project
- python launchers/pacman_gan_training.py --experiment=\<path to save the samples and models\>
- If you have GPUs, you can use --cuda to enable GPUs.

Generate fixed levels
- in the root folder of the project
- python launchers/pacman_gan_generate.py
    
Visualize the generated levels
- in the root folder of the project
- python launchers/pacman_grid_visualize.py

### End-to-end training
**Here the code only works with pytorch==1.2.0**
- in the root folder of the project
- python launchers/zelda_gan_partial_lp_end2end_generate.py --lvl_data=\<path to human authored levels>\ --gan_experiment=\<path to save the samples and models>\ --mipaal_experiment=\<path to save the mipaal samples and models>\

### Citing This Work
If you use this code for scholarly work, please kindly cite our work using the Bibtex snippet belw.
```
@inproceedings{zhang:aiide2020,
  title={Video Game Level Repair via Mixed Integer Linear Programming},
  author={Zhang, Hejia and Fontaine, Matthew and Hoover, Amy and Togelius, Julian and Dilkina, Bistra and Nikolaidis, Stefanos},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment},
  volume={16},
  number={1},
  pages={151--158},
  year={2020}
}
```

milp_constrained_gan

milp_constrained_gan 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

milp_constrained_gan

milp_constrained_gan 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README