/LeakGAN-PyTorch

A simple implementation of LeakGAN in PyTorch

Primary LanguagePython

LeakGAN-PyTorch

A simple implementation of LeakGAN in PyTorch described in Long Text Generation via Adversarial Training with Leaked Information.

Requirements

  • PyTorch r1.1.0
  • Python 3.5+
  • CUDA 8.0+ (For GPU)

File

  • Discriminator.py: The discriminator model of LeakGAN including Feature Extractor and classification
  • Generator.py: The generator model of LeakGAN including worker and manager units
  • data_iter.py: Data loader for Generator and Discriminator
  • utils.py: contains all the connecting parts for recurrent & loss functions
  • main.py: running this file will initiate
  • convert.py: Convert one-hot number to real word
  • eval_bleu.py: Evaluation of the BLEU scores (2-5) between test dataset and generated data

Reference

@article{guo2017long,
  title={Long Text Generation via Adversarial Training with Leaked Information},
  author={Guo, Jiaxian and Lu, Sidi and Cai, Han and Zhang, Weinan and Yu, Yong and Wang, Jun},
  journal={arXiv preprint arXiv:1709.08624},
  year={2017}
}

Acknowledgements

Main source:

  1. https://github.com/CR-Gjx/LeakGAN/blob/master/Image%20COCO/
  2. https://github.com/deep-art-project/Music/blob/master/leak_gan/

Copyright (c) 2019 Nurpeiis Baimukan