pretrain LayoutLM v2 on new dataset

[qifeishen]

Dear Authors, thanks for the great work. Do you plan to release the pretraining source code for LayoutLM v2?

[bonejay]

Yeah it would be a great help. Kind of tiresome that none of the document transformers (TILT, StructuralLM, layoutlm,StrucTexT) have release their pretraining code, although it is the most important factor to their performance.

[zzcgithub]

I hope so

[CheungZeeCn]

+1

[zzcgithub]

+2

[sudhirpol522]

+3

[rr191211]

I wonder when the source code will be released? Thanks

[ivsanro1]

Yeah it would be a great help. Kind of tiresome that none of the document transformers (TILT, StructuralLM, layoutlm,StrucTexT) have release their pretraining code, although it is the most important factor to their performance.

Just a guess, but they do this on purpose to force you to use their pre-trained models, and also to have control over their models and implementation. They never release pretraining codes (also for MarkupLM)