IBM/MAX-Image-Caption-Generator

Training on custom dataset ?

pidugusundeep opened this issue · 2 comments

Hey, I am curious if I could use my own dataset for training instead of the COCO dataset which is already being used.

Hello, we currently do not support retraining for this MAX model yet, but if you want to try it out on your own you can find the source code here.
https://github.com/tensorflow/models/tree/master/research/im2txt

Thanks will check that out @djalova