AI 中文总结
本研究提出PaDoc，一种基于布局的文档并行解码解析器，在OmniDocBench Full数据集上取得优异性能，且在A800 GPU上显著提升端到端文档解析的吞吐量并降低延迟。
AI 中文摘要
端到端文档解析器提供统一接口，但会将页面布局和区域内容序列化为自回归序列。这种表述方式强制将独立区域置于解码路径上，其长度随总内容量增长；而基于裁剪的两阶段解析器虽能实现区域级并行性，却需以重复视觉预填充和碎片化页面上下文为代价。为在保留全页面上下文的同时消除依赖，我们提出PaDoc，一种基于布局的解析器，将预测布局视为共享页面表示上的分支结构。在区域充分性假设下，我们推导了前缀条件分解，其中布局流和区域内容分支同步推进，将解码深度降低至最长布局-内容路径。我们在单个多模态大语言模型（MLLM）中实现了该分解：打包可变长度祖先注意力在标准下一个token训练下保持可见性，而掩码并行解码创建分支，经评估的vLLM后端将这些分支作为并发请求服务，利用缓存驻留的共享前缀复用。在OmniDocBench Full数据集上，PaDoc的整体布局F1值达91.1，在端到端解析器中获得顶级整体得分94.24，同时拥有最优文本编辑（0.038）和公式CDM（95.59）指标。在384页子集和单个A800 GPU上，它是5个并发级别下最快的端到端解析器，与同骨干的顺序SFT基线相比，有效页面吞吐量提升67.4%-118%，P95延迟降低39.2%-54.9%。代码可在该https URL获取。
英文摘要
End-to-end document parsers provide a unified interface, but serialize page layouts and regional contents into one autoregressive sequence. This formulation forces independent regions onto a decoding path whose length grows with the total content, whereas crop-based two-stage parsers expose region-level parallelism at the cost of repeated visual prefills and fragmented page context. To retain full-page context while removing dependencies, we propose PaDoc, a layout-grounded parser that treats the predicted layout as a branching structure over a shared page representation. Under a region-sufficiency assumption, we derive a prefix-conditioned factorization in which the layout stream and regional content branches advance concurrently, reducing the decoding depth to the longest layout-content path. We realize this factorization within a single MLLM: packed variable-length ancestor attention preserves the visibility under standard next-token training, while masked parallel decoding creates branches that the evaluated vLLM backend serves as concurrent requests with cache-resident shared-prefix reuse. On OmniDocBench Full, PaDoc attains an Overall layout F1 of 91.1 and, among end-to-end parsers, a top-tier Overall score of 94.24 together with the best Text Edit (0.038) and Formula CDM (95.59). On a 384-page subset and one A800 GPU, it is the fastest end-to-end parser at five concurrency levels, improving valid-page throughput by 67.4-118% and reducing P95 latency by 39.2-54.9% relative to a same-backbone Sequential SFT baseline. Code is available at https://github.com/Longin-Yu/Padoc
