MonlamAI/OCR-Tibetan_line_to_text_benchmark

这是一个用于评估和比较藏文OCR模型的线到文本基准数据集。数据集包含多种脚本、书写风格和打印方法，能够对多个领域的OCR模型进行全面测试。数据集被分为多个部分，每个部分代表不同来源的藏文文本，具有不同数量的示例和大小。

This is a line-to-text benchmark dataset for evaluating and comparing Tibetan OCR models. The dataset includes diverse scripts, writing styles, and print methods, enabling comprehensive testing of OCR models across multiple domains. The dataset is split into several subsets, each representing texts from various sources with different numbers of examples and sizes.