Skip to content

Dataset · lighteval

SIQA text dataset copy by Lighteval

lighteval/siqa

A text dataset containing between 10K and 100K records without a card description.

What's inside

Contains text data with a scale ranging from 10K to 100K examples. Card description is entirely absent, and neither language nor file format is specified.

What you can do with it

  • Evaluating text models via the Lighteval library
  • Basic text classification or generation experiments after manual structure verification
Size
10K–100K rows
Downloads this month
26.6K
Likes
10

How to load it

from datasets import load_dataset

ds = load_dataset("lighteval/siqa")