Roma Ushakov

@Axure did you find any solution?

By the way, I've discovered it works with cudnn 7.0.5 and tensorflow-gpu==1.8 from pip:)

Hi! I am not @Puzer, however, it seems he uses an output from the only one vgg layer

I would be very appreciated if you also could put "NN_test_pairs_trained_nw.txt" in this repo.

Thanks a lot! Could you please tell me which loss did you scored while training? And maybe you have any plot epochs vs. mse_loss on training? I mean loss on...