Training example

#1
by ysdj - opened

Hi,
Thanks a lot foe the trained model!

Any chance you can share the training code? I'm interested in how you prepared the DocVQA dataset for the fine tuning.
I tried following the instructions here:
https://github.com/PaddlePaddle/PaddleNLP/blob/develop/model_zoo/ernie-layout
but the model kind of overfitted.

Thanks!

Hi, could you share the training script / techniques you used? I can't seem to utilize this model to finetune on the SQuaD dataset. :)

Sign up or log in to comment