Prepare datasets for distributed training

You've preprocessed a dataset for a precision agriculture system to help farmers monitor crop health. Now you'll load the data by creating a DataLoader and place the data on GPUs for distributed training, if GPUs are available. Note the exercise actually uses a CPU, but the code is the same for CPUs and GPUs.

Some data has been pre-loaded:

A sample dataset with agricultural imagery
The Accelerator class from the accelerate library
The DataLoader class

This exercise is part of the course

Efficient AI Model Training with PyTorch

View Course

Exercise instructions

Create a dataloader for the pre-defined dataset.
Place the dataloader on available devices using the accelerator object.

Hands-on interactive exercise

Have a go at this exercise by completing this sample code.

accelerator = Accelerator()

# Create a dataloader for the pre-defined dataset
dataloader = ____(____, batch_size=32, shuffle=True)

# Place the dataloader on available devices
dataloader = accelerator.____(____)

print(accelerator.device)

Edit and Run Code