If you have any questions, feel free to ask! :)
To ensure the correctness of the experimental results, please run DCNv3 in FuxiCTR==2.0.1 or the latest version.

This model was formerly known as DCNv3

xDCN: Combining Exponential and Linear Cross Network for Click-Through Rate Prediction

Model Overview

Requirements

python>=3.6
pytorch>=1.10
fuxictr==2.0.1 or lastest
PyYAML
pandas
scikit-learn
numpy
h5py
tqdm

Experiment results

Datasets

Get the datasets from https://github.com/reczoo/Datasets

Hyperparameter settings and logs

Get the result from ./checkpoints

Acknowledgement

This implementation is based on FuxiCTR and BARS. Thanks for their sharing and contribution.
BARS: https://github.com/openbenchmark
FuxiCTR: https://github.com/xue-pai/FuxiCTR

Citation

If you find our code helpful for your research, please cite the following paper:

@article{li2024dcnv3,
  title={DCNv3: Towards Next Generation Deep Cross Network for CTR Prediction},
  author={Li, Honghao and Zhang, Yiwen and Zhang, Yi and Li, Hanwei and Sang, Lei and Zhu, Jieming},
  journal={arXiv preprint arXiv:2407.13349},
  year={2024}
}