DistilVDR：一種基於雙學生蒸餾的輕量級端到端視覺文檔檢索器

摘要

視覺文件檢索（VDR）主要由數十億參數的模型所主導，這些模型在完整語料庫規模下建置索引緩慢，且部署成本高昂。先前的壓縮方法要不是從零開始訓練較小的多向量編碼器，就是只對查詢端進行蒸餾；這兩種方法都無法端到端地產生緊湊的單向量檢索器。我們提出 DistilVDR，這是一個 524M 的端到端 VDR 系統，在逐點餘弦對齊損失下，從單一 8B 視覺-語言教師模型進行雙向蒸餾。所有監督皆來自凍結的教師模型嵌入空間，而該空間本身是透過相關性監督訓練的，因此學生模型的目標不需要相關性標籤、負採樣或對比項。我們以非對稱的僅編碼器學生模型來匹配 VDR 的文字查詢與圖像文件輸入不對稱性，將視覺容量集中在文件端，並將查詢端維持在 70M 參數。我們發布兩個變體，它們共享相同的編碼器和訓練，僅在文件編碼器的視覺區塊預算上有所不同：DistilVDR-HiRes 在 ViDoRe v1+v2+v3 上達到 61.74 的平均 NDCG@5（相當於 8B 教師模型的 86.9%），並在高解析度敏感的 v3 基準上領先所有已重現的低於 10 億參數基準模型；而 DistilVDR-Fast 在視覺令牌預算僅為前者的三分之一時達到 59.98。兩個變體將一百萬份文件儲存在比最強的低於 10 億參數多向量基準模型小 15.6 倍的索引中，並以快一個數量級的速度對語料庫進行索引。程式碼可於 https://github.com/Ryenhails/NanoVDR 取得。

English

Visual document retrieval (VDR) is dominated by multi-billion-parameter models that are slow to index at full corpus scale and expensive to serve. Prior compression routes either train a smaller multi-vector encoder from scratch or distil only the query side; neither yields a compact single-vector retriever end-to-end. We present DistilVDR, a 524M end-to-end VDR system distilled bilaterally from a single 8B vision-language teacher under a pointwise cosine alignment loss. All supervision comes from the frozen teacher's embedding space, which was itself trained with relevance supervision, so the student objective needs no relevance labels, negative sampling, or contrastive term. We match VDR's text-query and image-document input asymmetry with an asymmetric encoder-only student that concentrates visual capacity on the document side and keeps the query side at 70M parameters. We release two variants that share the same encoders and training and differ only in the document encoder's visual-tile budget: DistilVDR-HiRes attains 61.74 average NDCG@5 on ViDoRe v1+v2+v3 (86.9% of the 8B teacher) and leads every reproduced sub-1B baseline on the high-resolution-sensitive v3 benchmark, while DistilVDR-Fast attains 59.98 at a 3 times smaller visual-token budget. Both variants store one million documents in a 15.6 times smaller index than the strongest sub-1B multi-vector baseline and index the corpus an order of magnitude faster. The code is available at https://github.com/Ryenhails/NanoVDR.