---
license: cc-by-nc-4.0
---

<h1 align="center">
  PaDoc: Layout-Grounded Parallel Decoding for Document Parsing
</h1>

<p align="center">
  <a href="https://github.com/Longin-Yu/Padoc"><img src="https://img.shields.io/badge/GitHub-PaDoc-181717.svg?logo=github" alt="GitHub"></a>
  <a href="https://arxiv.org/abs/2608.06146"><img src="https://img.shields.io/badge/arXiv-2608.06146-b31b1b.svg" alt="arXiv"></a>
  <a href="https://huggingface.co/Longin-Yu/PaDoc"><img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Model-yellow" alt="Hugging Face"></a>
</p>

![Arch](main.png)

PaDoc is an open-source implementation of tree-supervised training and
fork-token decoding for document parsing. A single image-text model emits a
compact main sequence and starts independent content branches at learned fork
tokens. No draft model or additional prediction head is required.

For more details, refer to our github repo.