Closed dereyly closed 1 year ago
Hi @dereyly,
Thanks for your interest in the paper! The code is currently in a branch of a fork of EGG. It's not super documented as I'm planning to polish it and merge it before CVPR. In the meantime, you can find it here. Feel free to ask any questions you have until the code is merged.
Closing this, feel free to re open if you have any questions
Hi! @robertodessi, Where can I find the captions for your model finetuned on COCO and tested on COCO (Karpathy splits)?
Hi @NicholasMoratelli,
please send me an email at roberto.dessi11@gmail.com and I can send a set of captions. Apologies, but unfortunately I'm not sure these are the ones reported in the paper due to some lost access after I left meta, but I can sure we can check and work the numbers out. Otherwise it shouldn't be too hard or time-consuming to retrain a model checkpoint. Shoot me an email and can coordinate on this
Hi I read paper Cross-Domain Image Captioning with Discriminative Finetuning, it is very nice and can't find code https://github.com/facebookresearch/EGG/tree/main/egg/zoo/discriminative_captioner Do you release it?