This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
What is the difference between a PyTorch DataSet and a PyTorch DataLoader
A DataSet is designed to work with batches of data while a DataLoader is designed for retrieval of individual data items.
A DataSet is designed for retrieval of individual data items while a DataLoader is designed to work with batches of data.
The DataLoader class is the parent of the DataSet class
The DataSet class is the parent of the DataLoader class
Transforms in PyTorch are designed to:
perform some manipulation of the data to make it suitable for training
convert input tensors into an output tensor that contains a prediction
convert data items into visual representations
You must answer all questions before checking your work.
Need help? See our troubleshooting guide or provide specific feedback by reporting an issue.