Giters

X-PLUG / mPLUG-DocOwl

mPLUG-DocOwl: Modularized Multimodal Large Language Model for Document Understanding

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

如果是专注于文档理解任务的话，vision tower选用layoutlmv3初始化是不是会比vit更有竞争力？

whalefa1I opened this issue · comments

commented

如题

commented

Hi，@whalefa1I ， layoutlmv3是依赖ocr识别的文本和位置作为输入的，docowl系列都是不依赖ocr的～