There are no tokens_ id in sample，Could you explain it？

[ppzhr]

Traceback (most recent call last):
File "main.py", line 418, in
config.run(mode=args.mode)
File "main.py", line 161, in run
self.train()
File "main.py", line 404, in train
new_score, log = self.evaluation(dev_loader)
File "main.py", line 306, in evaluation
output = self.model(sample, is_train=False)
File "/home/jp/data/anaconda3/envs/openjere/lib/python3.8/site-packages/torch/nn/modules/module.py", line 727, in _call_impl
result = self.forward(*input, **kwargs)
File "/home/jp/zhr/copymtl/OpenJERE/openjere/models/seq2umt.py", line 112, in forward
result = self.decoder.test_forward(sample, o, h)
File "/home/jp/zhr/copymtl/OpenJERE/openjere/models/seq2umt.py", line 448, in test_forward
text_id = sample.tokens_id.cuda(self.gpu)
AttributeError: 'Batch_reader' object has no attribute 'tokens_id'