FLOPs metric

[guferreti]

How did you calculate the FLOPs of HDR-GAN, sorry for the out of topic question

[ltkong218]

I use the script like benchmark_SAFNet.py to test PyTorch implement of the generator of HDR-GAN.
Thanks.

[guferreti]

Can you share this implementation(I know that somethings you cannot share)? for my final paper(I'm undergraduating) and doing some modification in the generator model of HDR-GAN and I want to use FLOPs metric to compare the modification I did with the original(I used the original HDR-GAN repo for that).

[ltkong218]

The generator of HDR-GAN is a common U-Net and is easy to implement. Since you are doing some modification in the generator of HDR-GAN for your final paper, I suggest that you implement the code yourself for better understanding and more accurate FLOPs measurement.
Thanks.