KunpengLi1994 / VSRN

PyTorch code for ICCV'19 paper "Visual Semantic Reasoning for Image-Text Matching"
288 stars 47 forks source link

Do I need train two models for evaluation? #10

Closed weiyunfei closed 4 years ago

weiyunfei commented 4 years ago

Hi, I'm interested in your work, as well as appreciate for your code. However, I've not understood why it needs two pretrained models for evaluation. Should I train another model for evaluation? If so, is there any parameters that should be changed? I'll appreciate for your reply.

weiyunfei commented 4 years ago

Oops, sorry for this silly question. I found the solution in #5 just now and I will close this issue.