The "Rethinking" paper doesn't describe the actual implementation of the Inception-v3 model in Tensorflow: an accurate description is written in model.txt in the source files of the paper in the arXiv. Also the Inception-v2 implementation differs from the paper: referring to the figures in the paper, the implemented v2 is 2xfigure5 + 1xfigure10 + 4xfigure5 + figure10 + 2xfigure5. Does anybody know why it was implemented this way and not in a more faithful way?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.