| | --- |
| | license: apache-2.0 |
| | datasets: |
| | - gair-prox/open-web-math-pro |
| | language: |
| | - en |
| | base_model: |
| | - mistralai/Mistral-7B-v0.1 |
| | pipeline_tag: text-generation |
| | library_name: transformers |
| | --- |
| | |
| | # Mistral-7B-ProXMath |
| |
|
| | <p align="center"> |
| | <img src="prox-teaser.png"> |
| | </p> |
| |
|
| | [ArXiv](http://arxiv.org/abs/2409.17115) | [Data: OpenWebMath-Pro](https://huggingface.co/datasets/gair-prox/open-web-math-pro) | [Code](https://github.com/GAIR-NLP/program-every-example) |
| |
|
| | **Mistral-7B-ProXMath** is a math-adapted Mistral-7B-v0.1 model that is continually pre-trained on [OpenWebMath-Pro](https://huggingface.co/datasets/gair-prox/open-web-math-pro) (a refined version by ProX) for **10**B tokens. |
| |
|
| | ## Evaluations |
| |
|
| | ProX models are evaluated on 9 common math reasoning benchmarks. |
| |
|
| |
|
| | | Model | asdiv | gsm8k | mathqa | mawps | minerva_math | mmlu_stem | sat_math | svamp | tabmwp | average | |
| | |:---------------------:|:--------:|:--------:|:--------:|:--------:|:------------:|:---------:|:--------:|:--------:|:--------:|:--------:| |
| | | Mistral-7B-v0.1 | 68.5 | 40.6 | 32.3 | 87.0 | 11.4 | 50.0 | 56.2 | **65.4** | **52.9** | 51.6 | |
| | | Mistral-7B-ProXMath | **72.9** | **51.0** | **53.0** | **89.2** | **22.4** | **54.2** | **75.0** | 64.9 | 49.8 | **59.2** | |
| | |
| | |
| | |
| | ### 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} |
| | } |
| | ``` |
