Aan de slagGa gratis aan de slag

Manipulating datasets

There will likely be many occasions when you will need to manipulate a dataset before using it within a ML task. Two common manipulations are filtering and selecting (or slicing). Given the size of these datasets, Hugging Face leverages arrow file types.

This means performing manipulations are slightly different than what you might be used to. Fortunately, there's already methods to help with this!

The dataset is already loaded for you under wikipedia.

Deze oefening maakt deel uit van de cursus

Working with Hugging Face

Cursus bekijken

Praktische interactieve oefening

Probeer deze oefening eens door deze voorbeeldcode in te vullen.

# Filter the documents
____ = wikipedia.____(lambda row: "football" in row["____"])

# Create a sample dataset
example = ____.____(range(1))

print(example[0]["text"])
Code bewerken en uitvoeren