mmpretrain
mmpretrain copied to clipboard
BatchCutMixLayer和BatchMixupLayer中,prob并没有起到作用,所有数据100%被cutmix和mixup
https://github.com/open-mmlab/mmclassification/blob/d2e505415040bf5329ab218bb6fe3d899f176cd5/mmcls/models/utils/augment/mixup.py#L80
https://github.com/open-mmlab/mmclassification/blob/d2e505415040bf5329ab218bb6fe3d899f176cd5/mmcls/models/utils/augment/cutmix.py#L175
是不是应该有,如:
if random<self.prob:
return img, gt_label
else:
return self.cutmix(img, gt_label)
We recommend using English or English & Chinese for issues so that we could have broader discussion.
是这样的,这里的 prob 只是用作记录概率,随机选取是在 https://github.com/open-mmlab/mmclassification/blob/master/mmcls/models/utils/augment/augments.py 中进行的。 后续我们会对这块的逻辑进行重构
了解了,这里的prob代表选择Augments中某一个策略的概率