PaDoc：基於版面的並行解碼用於文件解析

摘要

端到端文檔解析器提供統一的介面，但將頁面佈局與區域內容序列化為單一自迴歸序列。此公式化迫使相互獨立的區域沿著一條解碼路徑依序生成，其長度隨總內容量增長；相較之下，基於裁剪的兩階段解析器雖能展現區域級並行性，卻需付出重複視覺預填充與頁面上下文碎片化的代價。為在保留完整頁面上下文的同時消除依賴，我們提出 PaDoc——一個以佈局為基礎的解析器，將預測出的佈局視為共享頁面表示之上的分支結構。在區域充分性假設下，我們推導出一種前綴條件分解，使佈局流與區域內容分支能並行推進，將解碼深度縮減至最長的佈局-內容路徑。我們在單一 MLLM 中實現此分解：打包的可變長度祖先注意力在標準下一詞元訓練下維持可見性，而遮罩並行解碼所建立的分支，由評測所用之 vLLM 後端以並發請求方式服務，並重用常駐快取中的共享前綴。在 OmniDocBench Full 上，PaDoc 的 Overall 佈局 F1 達到 91.1；在端到端解析器之中，Overall 分數達到頂尖的 94.24，同時取得最佳的 Text Edit（0.038）與 Formula CDM（95.59）。在 384 頁子集與單張 A800 GPU 設定下，PaDoc 在五種並發等級下均為最快的端到端解析器，相較於同骨幹的 Sequential SFT 基線，有效頁面吞吐量提升 67.4%–118%，P95 延遲降低 39.2%–54.9%。程式碼已公開於 https://github.com/Longin-Yu/Padoc。

English

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