bug-20220901

ValueError: Cannot load file containing pickled data when allow_pickle=False

pickle

feature = np.load(feature_url, allow_pickle=True)

 

OSError: Failed to interpret file 'array.npy' as a pickle

最后一个npy文件无用

self.feature = os.listdir(self.feature_dir)[:10]

 

Type 'MyDataset' doesn't have expected attribute '__iter__'

无法迭代，添加一条注释抑制警告，而不会完全禁用检查

1 # this comment is necessary to suppress an unnecessary PyCharm warning
2 # noinspection PyTypeChecker
3 train(args, model, device, train_dataset, optimizer, epoch)

 

RuntimeError: mat1 dim 1 must match mat2 dim 0

特征维度错误，Pytorch获取tensor维度

1 print(tag_feature.size()) #torch.Size([1, 200, 1, 1])
2 print(tag_feature.shape) #torch.Size([1, 200, 1, 1])

tag_feature = torch.tensor(tag_feature)

tag_feature = tag_feature.flatten() #拉平

 

RuntimeError: zero-dimensional tensor (at position 0) cannot be concatenated

两个数据不都是tensor类型