Preprocess a dataset for federated learning error TypeError: object.__new__() takes exactly one argument (the type to instantiate)

Question

我有一个 csv 数据集，我想将其分成 5 个部分，以将其分发给联邦学习场景中的 5 个客户端。这是我的代码：

train = train.iloc[:, 1:]

train = train.fillna(0)

train = train.rename(columns=lambda x: x.strip())

num_clients = 5

train_parts = np.array_split(train, num_clients)

train_data = []

for i in range(num_clients):
    # Convert the pandas dataframe into a PyTorch tensor
    data = torch.tensor(train_parts[i].iloc[:, :-1].values, dtype=torch.float32)
    labels = torch.tensor(train_parts[i].iloc[:, -1].values, dtype=torch.float32)

    scaler = MinMaxScaler()
    data = scaler.fit_transform(data.numpy())

    dataset = Dataset(data,labels)
    dataloader = DataLoader(dataset, batch_size=64, shuffle=True)
    train_data.append(dataloader)

我收到此错误消息：

<ipython-input-10-5b003ffb66f7> in <module>
     35     # data = torch.from_numpy(data).float()
     36 
---> 37     dataset = Dataset(data,labels)
     38     dataloader = DataLoader(dataset, batch_size=64, shuffle=True)
     39     train_data.append(dataloader)

/usr/lib/python3.8/typing.py in __new__(cls, *args, **kwds)
    873             obj = super().__new__(cls)
    874         else:
--> 875             obj = super().__new__(cls, *args, **kwds)
    876         return obj
    877 

TypeError: object.__new__() takes exactly one argument (the type to instantiate)

我的错误在哪里？我可以像这样拆分数据集并创建数据加载器吗？或者这是完全错误的？