[BUG] AssertionError: target None has to be real

[ZHUGUODONG1]

D:\Qgis\apps\Python312\Lib\site-packages\pytorch_lightning\utilities\parsing.py:269: UserWarning: Attribute 'loss' is an instance of nn.Module and is already saved during checkpointing. It is recommended to ignore them using self.save_hyperparameters(ignore=['loss']).
rank_zero_warn(
D:\Qgis\apps\Python312\Lib\site-packages\pytorch_lightning\utilities\parsing.py:269: UserWarning: Attribute 'logging_metrics' is an instance of nn.Module and is already saved during checkpointing. It is recommended to ignore them using self.save_hyperparameters(ignore=['logging_metrics']).
rank_zero_warn(
Traceback (most recent call last):
File "D:\我的下载\betterbatch-main\test.py", line 276, in
model = BatchDeepAREstimator()
^^^^^^^^^^^^^^^^^^^^^^
File "D:\我的下载\betterbatch-main\test.py", line 42, in init
super().init(**kwargs)
File "D:\我的下载\betterbatch-main\pytorch_forecasting\models\deepar_init_.py", line 140, in init
targeti in time_varying_reals_encoder
AssertionError: target None has to be real

[Sohaib-Ahmed21]

@ZHUGUODONG1 can you provide the reproducible code?

FYI , in pytorch-forecasting models, target must be given and should be real as expected in time series forecasting problems.