search

shabie / docformer

Implementation of DocFormer: End-to-End Transformer for Document Understanding, a multi-modal transformer based architecture for the task of Visual Document Understanding (VDU)
MIT License
253 stars 40 forks source link

Update dataset.py #3

Closed uakarsh closed 2 years ago

uakarsh commented 2 years ago

There is no SEP token appended