开始使用免费开始使用

Creating a train DataLoader

Now that we have split our dataset, we need to define a data loader to provide batches of data during training. DataLoader efficiently loads data into memory and allows shuffling for better generalization. In this exercise, you'll complete the train_dataloader method.

本练习是课程的一部分

Scalable AI Models with PyTorch Lightning

查看课程

练习说明

  • Import the DataLoader.
  • Return a DataLoader that loads self.train_data, enabling shuffling for better generalization.

交互式实操练习

通过完成这段示例代码来试试这个练习。

# Import libraries
from torch.utils.data import ____
import lightning.pytorch as pl

class LoaderDataModule(pl.LightningDataModule):
    def __init__(self):
        super().__init__()
        self.train_data = None
        self.val_data = None
    def setup(self, stage=None):
        self.train_data, self.val_data = random_split(dataset, [80, 20])
    def train_dataloader(self):
      	# Complete DataLoader
        return ____(____, batch_size=16, shuffle=____)
编辑并运行代码