SleepEarlyLiveLong

icon company Wuhan University, EIS icon location Wuchang District, Wuhan, Hubei P. icon location postgraduate student of Wuhan University, Electronic Engineering && Information and Communication Engineering

Results 5 comments of SleepEarlyLiveLong

multi-GPU for training

Thank you a lot! It works when I run "python main.py". However, when I run refine, it failed and gives error as follows: run: python main.py --refine --lr 1e-5 --reload...

multi-GPU for training

thank you! It is useful to use torch==1.7.1 to avoid that problem.

encounter errors when I try to finetune the model

> I meet the same error I solved the problem by updating the python from 3.9 to 3.10

Business Card & Receipt dataset

> The peper "OCR-free Document Understanding Transformer" uses two datasets _Business Card_ and _Receipt_. Can these two datasets be open source now? same question

I had conducted the experiment outlined in your paper and have come across results that do not match with the ones you reported.

hi, I encountered the same problem. I finetune the Ticket dataset using the same hyper-parameters provided by the code and get ted_accuracy=0.9454, f1_accuracy=0.8686. Do you solve the problem?