---
Model_Info:
   name: "ERNIE-Layout"
   description: 文心多语言跨模态布局增强文档智能大模型 
   description_en: ERNIE Cross-modal Pretrained Model
   update_time: 
   icon: "https://user-images.githubusercontent.com/11793384/203991610-5dbd3e53-4183-47ca-8d93-d99f8e22cc4d.png"
   from_repo: PaddleNLP
Task:
- tag: 自然语言处理
  tag_en: Natural Language Processing
  sub_tag: 文档分析
  sub_tag_en: Document Analysis
- tag: 文心大模型
  tag_en: Wenxin Big Models
  sub_tag: 文档分析
  sub_tag_en: Document Analysis
Example:
- tag: 互联网
  tag_en: Internet
  sub_tag: 文档智能
  sub_tag_en: Document Intelligence
  title: PaddleNLP文档智能技术重磅升级，动手搭建端到端文档抽取问答模型
  title_en: PaddleNLP document intelligent technology has been heavily upgraded, and the end-to-end document extraction question and answer model has been set up
  url: https://aistudio.baidu.com/aistudio/projectdetail/4881278?channelType=0&channel=0
  url_en: 
Datasets: FUNSD, XFUND-ZH, DocVQA-ZH, RVL-CDIP (sampled)
Publisher: Baidu
License: Apache 2.0
Paper: 
- title: "ERNIE-Layout: Layout-Knowledge Enhanced Multi-modal Pre-training for Document Understanding"
  url: https://arxiv.org/pdf/2210.06155.pdf
IfTraining: 1
IfOnlineDemo: 1