ShaitanRa/pascal-stahl-numarkdown

该数据集包含了使用NuMarkdown-8B-Thinking模型对[ShaitanRa/PascalStahl](https://huggingface.co/datasets/ShaitanRa/PascalStahl)数据集中的图像进行OCR处理后生成的markdown格式文本。处理详情包括源数据集、处理模型（numind/NuMarkdown-8B-Thinking）、样本数量（5个）、处理时间（4.1分钟）等。NuMarkdown-8B-Thinking是一个先进的基于推理的文档OCR模型，擅长处理复杂表格、数学公式、文档结构等，并生成格式良好的markdown输出。数据集结构包含所有原始列以及新增的markdown列（提取的markdown格式文本）和inference_info列（记录应用于该数据集的所有OCR模型的JSON列表）。

This dataset contains markdown-formatted OCR results from images in [ShaitanRa/PascalStahl](https://huggingface.co/datasets/ShaitanRa/PascalStahl) using NuMarkdown-8B-Thinking. Processing details include the source dataset, model (numind/NuMarkdown-8B-Thinking), number of samples (5), processing time (4.1 minutes), etc. NuMarkdown-8B-Thinking is a state-of-the-art reasoning-based document OCR model that excels at processing complex tables, mathematical formulas, document structures, etc., and generates well-formatted markdown output. The dataset structure contains all original columns plus markdown column (extracted text in markdown format) and inference_info column (JSON list tracking all OCR models applied to this dataset).