ValueError when computing AUC

[zeroduhyy]

During the training process, a ValueError occurs while computing the AUC in the metrics.py script. The error message indicates that at least two points are needed to compute the area under the curve, but only one point is available (x.shape = 1).

INFO:__main__:Training models (1/1)
INFO:simplenet:Training discriminator...
epoch:3 loss:0.39042 lr:0.0002 p_true:0.45 p_fake:0.482: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 4/4 [01:45<00:00, 26.32s/it]
...
...
...
ValueError: At least 2 points are needed to compute area under curve, but x.shape = 1  

The error suggests that the DataFrame df being passed to metrics.auc in metrics.py might not have enough valid data points for fpr and pro.
I don't know where i made mistakes, please teach me, thank you very much!