Questions on the implementation detail

[lykim200]

First of all, thank you for sharing this code! I am finding it really useful in trying to implement puzzle solver published by Noroozi, 2017. I have few questions on the implementation details.

1. Dataset/JigsawImageLoader.py line 55 what is the purpose of setting 0 standard deviation values to 1?

2. Noroozi mentioned in the original publication that "To prevent mapping the appearance to an absolute position we feed multiple Jigsaw puzzles of the same image to the CFN (an average of 69 out of 1000 possible puzzle configurations) and make sure that the tiles are shuffled as much as possible by choosing configurations with sufficiently large average Hamming distance" How is this being accomplished in your implementation? I understand that JigsawImageLoader outputs a single puzzle configuration per image. Do you simply run multiple epochs to ensure that training see multiple configurations per image?

3. Noroozi reports 59.5 hours of total training time (until convergence). How long did your implementation take to train until convergence?

[zgyang-hnu]

For question 3 mentioned above, given only one Titan X GPU was used for training (as mentioned in the paper), I am also wondering if it is possible to train over the ImageNet for 350K steps with batch size 256256256 for only 59.5 hours.