This quiz will test your knowledge of PyTorch DataLoaders. At the end you will be asked to provide an email where we will send you a copy of the results.
Click the button below to start.
Question 1 of 5
PyTorch DataLoaders perform which of the following functions?
Slices dataset into batches for training process
Speeds-up data loading with parallelization
Updates the model weights
Shuffles the data as it’s passed into the training process
Question 2 of 5
Setting num_workers to a value greater than 0 will:
Speed-up the data-loading process
Shuffle the data as the batches are created
Set the batch size for every mini-batch
Question 3 of 5
You will need to create a separate DataLoaders for the training, testing and validation datasets.
True
False
Question 4 of 5
Which setting will speed-up data loading if using a CUDA device by removing a memory copy:
num_workers
pin_memory
drop_last
Question 5 of 5
If you want to avoid using the last mini-batch during training if your total dataset is not divisible by the batch_size you would:
Set num_workers to 0
Set shuffle to True
Set drop_last to True
Set pin_memory to True