BEE-spoke-data/napierone-pdf-olmOCR
收藏Hugging Face2025-12-29 更新2025-04-12 收录
下载链接:
https://hf-mirror.com/datasets/BEE-spoke-data/napierone-pdf-olmOCR
下载链接
链接失效反馈官方服务:
资源简介:
NapierOne PDFs数据集是通过olmOCR管道将NapierOne的PDF文档转换为文本的数据集。该数据集过滤掉了输出小于等于100个字符和低质量的PDF文档。在默认配置中,4228行数据大约代表了93,595个输入PDF页面。
The NapierOne PDFs dataset is a collection of text converted from NapierOne PDF documents using the olmOCR pipeline. The dataset filters out documents with outputs of 100 characters or less and those of low quality. In the default configuration, 4228 rows represent approximately 93,595 input PDF pages.
提供机构:
BEE-spoke-data



