The problem of reproducing the result of ImageNet about adabelief-optimizer HOT 4 OPEN

Hi, the original code I used for ImageNet is in the example folder, but should be the same as 0.2.0, except that 0.2.0 uses a default eps=1e-16, the example used 1e-8, but I think it is override to 1e-8 explicitly (rather than using default) in the script.

from adabelief-optimizer.

Hi, the original code I used for ImageNet is in the example folder, but should be the same as 0.2.0, except that 0.2.0 uses a default eps=1e-16, the example used 1e-8, but I think it is override to 1e-8 explicitly (rather than using default) in the script.

Thank you for your reply. I also think it has been overrided to 1e-8. So do you think the edition of the PyTorch could make a difference to the result? And mine is 1.11.0.

from adabelief-optimizer.

I'm not quite sure, pytorch version might case slight difference. BTW, I remember there are different versions of ImageNet, after 2018 or some year some of the data is re-labeled or replaced. Did you use the latest version with corrections?

from adabelief-optimizer.

I used the ILSVRC2012 edition of ImageNet.

from adabelief-optimizer.