HPD-Parsing：階層式並行文件解析

摘要

高效的團隊協作通常結合全域協調與並行執行，然而這一原則尚未在基於統一視覺語言模型（VLM）的文件解析器中得到充分體現。現有的統一解析器雖能共同處理整頁內容，但其輸出過程卻依賴單一的逐詞自回歸生成軌跡，這形成了隨文件長度增長的序列化瓶頸。這種全頁序列生成忽略了文件解析的關鍵特性：版面結構需全域分析，而區塊內容則可並行解析。基於此觀察，我們提出HPD-Parsing，以分層並行解碼（Hierarchical Parallel Decoding）典範取代全頁自回歸生成。主版面分支負責組織整體文件結構，並動態分配區塊級內容解碼至並行分支，同時漸進式多詞預測（P-MTP）進一步減少各分支內的解碼步驟。在公開基準測試中，HPD-Parsing每秒可處理4,752個詞元，吞吐量達到現有最快文件解析模型的2.62倍，以及傳統自回歸基線模型的3.06倍，同時維持具競爭力的解析準確率。這些結果證明了分層並行解碼可作為全頁自回歸生成的有效替代方案，為高效統一文件解析開闢了新方向。

English

Efficient teamwork typically combines global coordination with parallel execution, a principle not yet fully reflected in unified Vision-Language Model (VLM)-based document parsers. Existing unified parsers process an entire page jointly but generate its output through a single token-by-token autoregressive trajectory, creating a sequential bottleneck that grows with document length. Such full-page sequential generation overlooks a key property of document parsing: layout must be analyzed globally, whereas block content can be parsed in parallel. Based on this observation, we introduce HPD-Parsing, which replaces full-page autoregressive generation with a Hierarchical Parallel Decoding paradigm. A main layout branch organizes the overall document structure and dynamically assigns block-level content decoding to concurrent branches, while progressive multi-token prediction (P-MTP) further reduces the decoding steps within each branch. Experiments on public benchmarks show that HPD-Parsing achieves 4,752 tokens per second, delivering 2.62times the throughput of the fastest existing document parsing model and 3.06times that of the vanilla autoregressive baseline, while maintaining competitive parsing accuracy. These results establish hierarchical parallel decoding as an effective alternative to full-page autoregressive generation, opening a new direction for efficient unified document parsing.