search

juntang-zhuang / Adabelief-Optimizer

Repository for NeurIPS 2020 Spotlight "AdaBelief Optimizer: Adapting stepsizes by the belief in observed gradients"
BSD 2-Clause "Simplified" License
1.05k stars 108 forks source link

On imagenet accuracy result 70.08 #50

Closed wyzjack closed 3 years ago

wyzjack commented 3 years ago

Hi, congrats on the nice work. But I have a problem in achieving your claimed accuracy result 70.08 of the ImageNet experiment in the paper. My run on my machine using your code with your parameter setting is 69.32. image Could you please provide (link to) your model checkpoint file, or is there any other tricks in training? Thanks .

juntang-zhuang commented 3 years ago

Hi, I don't have pretrained weights available now, too long ago. If I remember correctly, there might be some wrongly labled validation images in ImageNet12, and you will need to "blacklist" those images. That might cause the difference. Also random seed might be a reason.