xcyao00/BGAD

训练过程中,PIX_AUPRO 为-100

Closed this issue · 1 comments

DONTKK commented

你好,我在mvtec数据集上训练,没有更改代码,训练结果的 PIX_AUPRO 都是-100正常吗

Epoch: 24 test_loss: 2.9807
IMG_AUROC: last: 100.00 max: 100.00 epoch_max: 4
PIX_AUROC: last: 99.64 max: 99.67 epoch_max: 21
PIX_AUPRO: last: -100.00 max: -100.00 epoch_max: 0
Saving weights to output/bgad_fas_10/weights/mvtec_tf_efficientnet_b6_conditional_flow_model_hazelnut.pt

It's ok. As the pix_aupro calculation is time-consuming, I just set this to -1 when training. You can save the best checkpoint and calculate the pix_aupro for this checkpoint. Or you can change the code to calculate the pix_aupro every epoch.