Giters

avirambh / MSDNet-GCN

ICLR 2018 reproducibility challenge - Multi-Scale Dense Convolutional Networks for Efficient Prediction

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Question about Dynamic Evaluation

YihengJiang opened this issue · comments

commented

Hi author,
thanks for your work, but I have not understand the dynamic evaluation in origin paper, in this case, i can not implement early exiting from paper.
Can you send me the code about this point? cause I think I can understand paper by code. Thank you very much.

commented

Hi @YihengJiang,
Sorry for the late reply.
I think there is an open issue about dynamic evaluation / early exiting in the original Lua repository, where they explain how they implement it (finding a confidence threshold of the validation accuracy). Of course you are very welcome to contribute an example for this approach.
Aviram