The pretrained STME module

[imyzx2017]

Did you re-trained the STME module based on 'attnGAN' ? If so, could you provided it? thanks!

After adding my own pretrained caption model, replacing the DAMSM loss to caption_loss doesn't improve the IS, can you give some advices?

[mkumar10]

I think they add it with the DAMSM loss not replace it. Then you should be able to reproduce.

[saramahersalem]

@imyzx2017 Hello @imyzx2017 , could you help me to train model to my own caption dataset?

[rfelixmg]

@qiaott in the paper you have experimented on CUB-Birds. However, in this repo you have added STREAM model only for COCO. Can you possibility release the weights for CUB-BIRDS data set as well?

Best wishes, Rafa.