About the eval_model.py

[PanXiongAdam]

Hello, I ran the eval_model.py file to test the training model. The best performance of this model is 0.65 in training, but I found that the running result on eval_model is 0.11. Have you tested eval_ model?

[baihuple]

Hello sir, when I use the link you provided about refer, its' mscoco data link is out of date so i cannot form the right data. And when debugging, the result turned out that ValueError: too many values to unpack (expected 3) which is from "img_file, ref_id, phrase = self.images[idx]". If you can give me some help, i'll be very grateful!

[PanXiongAdam]

@baihuple Sorry, I never solve it. I currently concern other work.

[baihuple]

@PanXiongAdam Thanks for your commit