运行出错

[11140321]

我想问一下，为什么运行完一个epoch之后会出现这样的错误？
epoch:0 train_rmse:7.3017 valid_rmse:6.7257 time:48.230 s
Traceback (most recent call last):
File "D:/DT/InteractionGraphNet-master/codes/ign_train.py", line 184, in
valid_true, valid_pred, _ = run_a_eval_epoch(DTIModel, valid_dataloader, device)
File "D:/DT/InteractionGraphNet-master/codes/ign_train.py", line 54, in run_a_eval_epoch
outputs = model(bg, bg3)
File "F:\anaconda1\envs\torchcpu\lib\site-packages\torch\nn\modules\module.py", line 541, in call
result = self.forward(*input, **kwargs)
File "D:\DT\InteractionGraphNet-master\codes\model.py", line 409, in forward
atom_feats = bg.ndata.pop('h')
File "F:\anaconda1\envs\torchcpu\lib_collections_abc.py", line 795, in pop
value = self[key]
File "F:\anaconda1\envs\torchcpu\lib\site-packages\dgl\view.py", line 60, in getitem
return self._graph.get_n_repr(self._nodes)[key]
KeyError: 'h'

[zjujdj]

@11140321
我感觉可能是你的数据有问题？或者你调调batch_size看看？