gair-prox/math-doc-refining-lm

Math-doc-refining-lm

<p align="center"> <img src="prox-teaser.png"> </p>

ArXiv | Code

Math-doc-refining-lm is an adapted 0.7B-ProX model, fine-tuned for doc level refining via program generation, and can be applied over math pre-training corpus such as open-web-math.

<p align="center"> <img src="func_design.png"> </p>

Citation

@article{zhou2024programming,
  title={Programming Every Example: Lifting Pre-training Data Quality like Experts at Scale},
  author={Zhou, Fan and Wang, Zengzhi and Liu, Qian and Li, Junlong and Liu, Pengfei},
  journal={arXiv preprint arXiv:2409.17115},
  year={2024}
}