Is it normal that a simple split on the mrqa dataset takes this long (~ 45 mins)? Or is colab pro slow?
This is the code
from datasets import load_dataset mrqa = load_dataset( 'mrqa', split='train[:10000]' )
