Gsunshine / phantom_grad

[NeurIPS 2021] On Training Implicit Models

On Training Implicit Models

This is the official repo for paper, On Training Implicit Models (NeurIPS 2021).

Organization

You may expect to train your own implicit models through phantom gradients and reproduce the results in this paper. We offer a speedup of about 2 times regarding training DEQs.

• MDEQ
• DEQ
• IGNN

Under construction. More components will be added later.

Reference

If you find our work helpful to your research, please consider citing this paper. :)

Bibtex is here.

@article{geng2021training,
  title={On Training Implicit Models},
  author={Geng, Zhengyang and Zhang, Xin-Yu and Bai, Shaojie and Wang, Yisen and Lin, Zhouchen},
  journal={Advances in Neural Information Processing Systems},
  year={2021}
}

Contact

Feel free to contact me if you have additional questions or have interests in collaboration. Please drop me an email at [e-mail]. Find me at Twitter or WeChat. Thank you!