Kaze816
Kaze816
> 可以换成 paddleseg 2.9试一下 2.9的问题更多，我都没有跑到这里来
because it need 10Gb+ memory when training by one
> what if batch size > 1 in the train process? maybe out of memory size.
long time to wait for
> > long time to wait for > > I advise you to give up waiting！ Maybe, you are right.
> I hypothesize that InvAD likely relies on operators in `model/stylegan2_op`, and these operators are non-native implementations by PyTorch. You can search to see if anyone has successfully exported StyleGANv2...
配置文件中的tain.txt 中是什么样式呀？ @alanzheng800 这不是只需要正样本么，难道也需要生成些0值的标签图像？ train_dataset: type: Dataset num_classes: 1 dataset_root: /mv_dataset/grid # You need to specify a specific category root in mvtec_ad dataset **train_path: mv_dataset/grid/train.txt** ...