features 字典缺少 label ，keyerror问题

[ZuoMM-f]

在运行MKG任务bash scripts/pretrain_fb15k-237-image.sh脚本时候出现的缺少label的问题，请教这个问题的解决方法，wn18数据集也有这个问题，数据集下载的是作者提供的[GoogleDrive]
Epoch 0: 80%|███████████████████████████████████████████████████████████████████████████████████▉ | 935/1169 [48:27<12:07, 3.11s/it, loss=11.2, v_num=9, hits10=0.000, hits1=0.000]
Traceback (most recent call last): | 0/234 [00:00<?, ?it/s]
File "main.py", line 153, in
main()
File "main.py", line 135, in main
trainer.fit(lit_model, datamodule=data)
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 458, in fit
self._run(model)
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 756, in _run
self.dispatch()
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 797, in dispatch
self.accelerator.start_training(self)
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/pytorch_lightning/accelerators/accelerator.py", line 96, in start_training
self.training_type_plugin.start_training(trainer)
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/pytorch_lightning/plugins/training_type/training_type_plugin.py", line 144, in start_training
self._results = trainer.run_stage()
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 807, in run_stage
return self.run_train()
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 869, in run_train
self.train_loop.run_training_epoch()
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/pytorch_lightning/trainer/training_loop.py", line 576, in run_training_epoch
self.trainer.run_evaluation(on_epoch=True)
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py", line 949, in run_evaluation
for batch_idx, batch in enumerate(dataloader):
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 630, in next
data = self._next_data()
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/torch/utils/data/dataloader.py", line 673, in _next_data
data = self._dataset_fetcher.fetch(index) # may raise StopIteration
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/torch/utils/data/_utils/fetch.py", line 55, in fetch
return self.collate_fn(data)
File "/home/zsy/MKGformer/MKG/data/data_module.py", line 103, in call
label = [feature.pop("label") for feature in features]
File "/home/zsy/MKGformer/MKG/data/data_module.py", line 103, in
label = [feature.pop("label") for feature in features]
KeyError: 'label'
Exception ignored in: <function tqdm.del at 0x73a1b43b2430>
Traceback (most recent call last):
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/tqdm/std.py", line 1090, in del
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/tqdm/std.py", line 1303, in close
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/tqdm/std.py", line 1481, in display
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/tqdm/std.py", line 1093, in repr
File "/opt/conda/envs/MKGformer/lib/python3.8/site-packages/tqdm/std.py", line 1443, in format_dict
TypeError: cannot unpack non-iterable NoneType object

[njcx-ai]

您好，建议您检测下看看下载后的数据集读取后是否包含“label”字段呢

[ZuoMM-f]

{'input_ids': [101, 1996, 6412, 1997, 103, 2003, 2008, 1037, 2223, 1997, 2037, 2219, 2003, 1037, 2826, 2137, 4038, 1011, 3689, 2143, 2008, 4136, 1037, 7214, 3550, 4070, 1997, 1996, 2613, 1011, 2166, 2035, 1011, 2137, 3057, 2658, 3598, 2223, 1012, 2856, 2011, 10647, 5832, 1010, 1996, 2143, 3340, 3419, 9180, 2015, 1010, 20277, 2532, 4482, 1010, 15820, 1051, 1005, 18016, 1010, 1998, 11284, 1012, 102], 'attention_mask': [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], 'label': 9929, 'entity': '/m/02ndy4'}
Epoch 0: 80%|███████████████████████████████████████████████████████████████████████████████████▏ | 935/1169 [47:44<11:56, 3.06s/it, loss=11.2, v_num=10, hits10=0.000, hits1=0.000]
{'input_ids': [101, 1996, 6412, 1997, 103, 2003, 2008, 21173, 1011, 18864, 2003, 1996, 22037, 1997, 1996, 7032, 2163, 1997, 2762, 1010, 9605, 2087, 1997, 1996, 3439, 11068, 1997, 18864, 1998, 1996, 2670, 2112, 1997, 1996, 2280, 11068, 1997, 21173, 1012, 2049, 3007, 2103, 2003, 20963, 1025, 2060, 3862, 3655, 2024, 11320, 12750, 1010, 13109, 28812, 1998, 11265, 2819, 4609, 6238, 1012, 1996, 2280, 2394, 102], 'attention_mask': [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]}
没有找到 'label' 键 {'input_ids': [101, 1996, 6412, 1997, 103, 2003, 2008, 21173, 1011, 18864, 2003, 1996, 22037, 1997, 1996, 7032, 2163, 1997, 2762, 1010, 9605, 2087, 1997, 1996, 3439, 11068, 1997, 181998, 1996, 2670, 2112, 1997, 1996, 2280, 11068, 1997, 21173, 1012, 2049, 3007, 2103, 2003, 20963, 1025, 2060, 3862, 3655, 2024, 11320, 12750, 1010, 13109, 28812, 1998, 11265, 2819, 4609, 6238, 1012, 1996, 2280, 2394, 102], 'attention_mask': [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]}

label值貌似是生成的一个数，不是所有的entity都没有label值，打印出来之后发现大部分有生成的label值，后面的没有，但是没有label的好像也没有entity，不知道这段代码是否有问题

[ZuoMM-f]

有没有可能是data_module文件中，关于label列表没有对缺少“label”key的处理导致的问题，"labels"，“entity”都有处理。

或者需要对缺少图片实体添加文件夹这样的处理吗？恳请作者回答我的问题，非常感谢

[njcx-ai]

您好，我记得应该是有需要对缺少图片实体添加文件夹这样的处理的，data_module中之前我们关于“label”key也都是能跑通的呀。

[zxlzr]

请问您的问题解决了吗