search

woojeongjin / FewVLM

A Good Prompt Is Worth Millions of Parameters: Low-resource Prompt-based Learning for Vision-Language Models (ACL 2022)
40 stars 5 forks source link

Some files for VQA v2 may be missing.. #3

Open yuanrr opened 1 year ago

yuanrr commented 1 year ago

When I trian on VQA v2 by default:

CUDA_VISIBLE_DEVICES=$1 python src/vqa.py \ --train train \ --valid train \ --test minival,nominival \

It reported: No such file or directory: './train.json' I noticed that VQA dataset provided by VLT5 did not have those files. Am I mistaken here... Hope to get your response, thanks!

yuanrr commented 1 year ago

Sorry for disturbing you, I find it in LXMERT. : )