olmOCR 2：文件光學字符識別單元測試獎勵

摘要

我们隆重推出olmOCR 2，这是我们家族中最新一代强大的光学字符识别（OCR）系统，专为将数字化印刷文档（如PDF）转换为整洁、自然顺序的纯文本而设计。olmOCR 2的核心动力源自olmOCR-2-7B-1025，这是一款经过强化学习与可验证奖励（RLVR）训练的专业化7B视觉语言模型（VLM），其中我们的奖励机制基于一系列多样化的二元单元测试。为了扩大单元测试的创建规模，我们开发了一套流程，用于生成具有多样性和挑战性布局的合成文档，这些文档附有已知的真实HTML源代码及提取的测试案例。我们证明，在这些测试案例上进行RL训练，使得olmOCR 2在olmOCR-Bench——我们的英语OCR基准测试中，达到了业界领先的性能，尤其是在数学公式转换、表格解析以及多栏布局处理方面，相较于前代版本实现了最大幅度的提升。我们以宽松的开源许可发布了我们的模型、数据及代码。

English

We present olmOCR 2, the latest in our family of powerful OCR systems for converting digitized print documents, like PDFs, into clean, naturally ordered plain text. olmOCR 2 is powered by olmOCR-2-7B-1025, a specialized, 7B vision language model (VLM) trained using reinforcement learning with verifiable rewards (RLVR), where our rewards are a diverse set of binary unit tests. To scale unit test creation, we develop a pipeline for generating synthetic documents with diverse and challenging layouts, known ground-truth HTML source code, and extracted test cases. We show that RL training on these test cases results in state-of-the-art performance on olmOCR-Bench, our English-language OCR benchmark, with the largest improvements in math formula conversion, table parsing, and multi-column layouts compared to previous versions. We release our model, data and code under permissive open licenses.

olmOCR 2：文件光學字符識別單元測試獎勵

olmOCR 2: Unit Test Rewards for Document OCR

摘要

Support