tanshaosuannai

I am currently working with the train_faster_rcnn_alt_opt.py script from the Bottom-Up-Attention project. However, I noticed that the required model files stage1_rpn_solver60k80k.pt and stage1_fast_rcnn_solver30k40k.pt do not appear to be available in...

大佬，我在用mukea训练自己的数据集时报错，没有MuKEA/data/kr-vqa/krvqa-ans_dic.pickle这个文件，我看您readme中也没有写，这个文件是如何色生成的，有什么用啊？