aurooj / WSG-VQA-VLTransformers

Weakly Supervised Grounding for VQA in Vision-Language Transformers
MIT License
16 stars 2 forks source link

Difference between lxmert_pretrain.py and lxmert_pretrain_spatial.py #4

Closed zxzhou9 closed 1 year ago

zxzhou9 commented 1 year ago

Sorry to bother you again, when I studied the code, I found the two file. While in the bash code, you almost use lxmert_pretrain.py for all, so what's the difference between them?

Second problem is due to the limit of my server, I can only use batchsize:8, thus I need to run 20 hours for one epoch, do I have any method to accelerate it?