README.md

RL-Chord

This is the official implementation of RL-Chord, which aims to generate chord progressions for the given melody using reinforcement learning (RL) techniques.

• Paper link
• Check our demo page and listen!
  

Data

Wiki training data can be downloaded here.

CLSTM

RL-Chord

DQN-Chord

Reference

If you find the code useful for your research, please consider citing

@article{ji2023rl,
  title={RL-Chord: CLSTM-Based Melody Harmonization Using Deep Reinforcement Learning},
  author={Ji, Shulei and Yang, Xinyu and Luo, Jing and Li, Juan},
  journal={IEEE Transactions on Neural Networks and Learning Systems},
  year={2023},
  publisher={IEEE}
}

Acknowledgments

We appreciate RL-Adventure-2 for providing the reference codes of various reinforcement learning algorithms.