Pytorch 使用报错 RuntimeError: Caught RuntimeError in DataLoader worker process 0.

Mister Leon2025-07-17 18:24

这个错误是可能是由于在DataLoader的工作进程中尝试访问CUDA设备导致的。PyTorch的DataLoader使用多进程加载数据，而CUDA上下文不能在子进程中直接使用。

修改前的代码为：

复制代码

def prepare_data(file_path):
    # 读取Excel文件
    df = pd.read_excel(file_path, header=None)
    df = df.iloc[1:]
    print(df)

    # 提取特征和标签
    features = df.iloc[:, :-1].values.astype('float32')  # extract feature
    labels = df.iloc[:, -1].values.astype('int64')  # extract label



    # 数据标准化
    scaler = StandardScaler()
    features = scaler.fit_transform(features)

    # 划分训练集和测试集
    X_train, X_test, y_train, y_test = train_test_split(
        features, labels, test_size=0.2, random_state=42
    )

    # 转换为PyTorch张量并移动到设备
    X_train = torch.tensor(X_train, device=device)
    X_test = torch.tensor(X_test, device=device)
    y_train = torch.tensor(y_train, device=device)
    y_test = torch.tensor(y_test, device=device)

    return X_train, X_test, y_train, y_test, scaler

数据加载修改为下运行OK：

复制代码

class ExcelDataset(Dataset):
    def __init__(self, features, labels):
        # 确保数据在CPU上
        self.features = features.cpu() if features.is_cuda else features
        self.labels = labels.cpu() if labels.is_cuda else labels