Justherozen/ProMix
[IJCAI 2023] ProMix: Combating Label Noise via Maximizing Clean Sample Utility
PythonMIT
Issues
- 0
Hyperparameter lr_decay_epochs missing problem
#16 opened by Pallaksch - 0
- 2
About inference and train
#13 opened by nengwp - 7
warmup performance good but after poor
#14 opened by PalaceTony - 1
warmup training
#12 opened by nengwp - 2
CSS samples num in imbalanced dataset
#11 opened by nengwp - 2
loss: nan
#7 opened by youngjae0129 - 2
strong or weak augmentation have an ambiguity
#10 opened by nengwp - 1
- 0
About lr in train
#8 opened by wangnaihao - 11
您觉得这个方法具备通用性吗?对于不同的分类任务/不同的分类模型而言。
#2 opened by Twsaiae - 1
At cifar10 IDN noise rate setting, cann't repeat the results reported at paper.
#6 opened by zhengzhang37 - 3
- 1
Reproducing the experiment
#5 opened by ghoshsdeeep - 2
- 0