yuyq96/R1-Vision-PixMo-Cap-QA
收藏Hugging Face2025-02-08 更新2025-02-15 收录
下载链接：
https://hf-mirror.com/datasets/yuyq96/R1-Vision-PixMo-Cap-QA
下载链接
链接失效反馈官方服务：
资源简介：
R1-Vision项目使用的数据集包括文本数据、文本渲染数据和多模态数据。文本数据来自Bespoke-Stratos-17k数据集，文本渲染数据同样来自Bespoke-Stratos-17k，经过重格式化和图像渲染处理。多模态数据来自AI2D、ScienceQA和PixMo-Cap-QA数据集。这些数据集用于训练一个能够处理文本和图像的双模态模型。
The R1-Vision project uses datasets that include text data, text rendering data, and multimodal data. The text data is from the Bespoke-Stratos-17k dataset, the text rendering data is also from Bespoke-Stratos-17k after reformatting and image rendering, and the multimodal data is from the AI2D, ScienceQA, and PixMo-Cap-QA datasets. These datasets are used to train a bimodal model capable of processing both text and images.
提供机构：
yuyq96
