ImgCode-8.6M
收藏资源简介：
ImgCode-8.6M 是目前最大的图像-代码数据集，由香港中文大学多媒体实验室（MMLab）的研究团队开发。该数据集通过图像到代码的模型 FigCodifier 生成，能够将数学相关的图像转换为详细的代码，进而渲染新的图像。ImgCode-8.6M 包含了 860 万个图像-代码对，这些数据对通过迭代训练的方式生成，确保了数据的准确性和多样性，为大型多模态模型的数学推理能力提供了强有力的支持。
ImgCode-8.6M is currently the largest image-code dataset, developed by the research team of the Multimedia Laboratory (MMLab) at The Chinese University of Hong Kong. This dataset is generated via the image-to-code model FigCodifier, which can convert math-related images into detailed code that can render new images. ImgCode-8.6M contains 8.6 million image-code pairs, which are generated through iterative training to ensure the accuracy and diversity of the data, providing robust support for the mathematical reasoning capabilities of large multimodal models.
MathCoder数据集概述
数据集基本信息
- 名称: MathCoder
- 相关论文:
数据集和模型
- 数据集:
- 模型:
- Base Model: Llama-2:
- Base Model: Code Llama:
- MathCoder-VL Models:
训练数据
- 训练数据集: MathCodeInstruct
- 数据特点: 每个解决方案交织了自然语言、代码和执行结果。
方法介绍
- 方法: 生成新颖且高质量的数据集，包含数学问题及其基于代码的解决方案。
- 目标: 通过代码建模和推导数学方程，增强语言模型的数学推理能力。
性能表现
- MATH数据集: 45.2%
- GSM8K数据集: 83.9%
- 其他成就:
- 在GSM8K和MATH数据集上超越ChatGPT-3.5和PaLM-2。
- 在竞赛级MATH数据集上超越GPT-4。
使用方式
- 模型部署: 使用Text Generation Inference (TGI)工具包部署。
- 推理: 提供推理脚本，支持自定义IP和端口。
- 评估: 提供评估脚本，用于评估预测答案。
引用
bibtex @inproceedings{ wang2025mathcodervl, title={MathCoder-{VL}: Bridging Vision and Code for Enhanced Multimodal Mathematical Reasoning}, author={Ke Wang and Junting Pan and Linda Wei and Aojun Zhou and Weikang Shi and Zimu Lu and Han Xiao and Yunqiao Yang and Houxing Ren and Mingjie Zhan and Hongsheng Li}, booktitle={The 63rd Annual Meeting of the Association for Computational Linguistics}, year={2025}, url={https://openreview.net/forum?id=nuvtX1imAb} }
bibtex @inproceedings{ lu2025mathcoder2, title={MathCoder2: Better Math Reasoning from Continued Pretraining on Model-translated Mathematical Code}, author={Zimu Lu and Aojun Zhou and Ke Wang and Houxing Ren and Weikang Shi and Junting Pan and Mingjie Zhan and Hongsheng Li}, booktitle={The Thirteenth International Conference on Learning Representations}, year={2025}, url={https://openreview.net/forum?id=1Iuw1jcIrf} }
bibtex @inproceedings{ wang2024mathcoder, title={MathCoder: Seamless Code Integration in {LLM}s for Enhanced Mathematical Reasoning}, author={Ke Wang and Houxing Ren and Aojun Zhou and Zimu Lu and Sichun Luo and Weikang Shi and Renrui Zhang and Linqi Song and Mingjie Zhan and Hongsheng Li}, booktitle={The Twelfth International Conference on Learning Representations}, year={2024}, url={https://openreview.net/forum?id=z8TW0ttBPp} }
bibtex @inproceedings{ zhou2024solving, title={Solving Challenging Math Word Problems Using {GPT}-4 Code Interpreter with Code-based Self-Verification}, author={Aojun Zhou and Ke Wang and Zimu Lu and Weikang Shi and Sichun Luo and Zipeng Qin and Shaoqing Lu and Anya Jia and Linqi Song and Mingjie Zhan and Hongsheng Li}, booktitle={The Twelfth International Conference on Learning Representations}, year={2024}, url={https://openreview.net/forum?id=c8McWs4Av0} }
