Skip to content

Dataset · criteo

Criteo 1TB ad click logs

criteo/CriteoClickLogs

Massive tabular dataset with billions of rows containing ad features and click labels for CTR prediction models.

What's inside

The dataset contains feature values and click feedback for millions of display advertising impressions, split into 24 files representing one day of data each. It totals between 1 and 10 billion records in tabular and text formats. It resembles the well-known Kaggle Criteo Display Advertising Challenge dataset but exceeds it in scale.

What you can do with it

  • Training and benchmarking click-through rate (CTR) prediction algorithms
  • Testing scalable recommender systems and models for large-scale tabular data
Tasks
Text classification
Size
1B–10B rows
Downloads this month
42.3K
Likes
15

How to load it

from datasets import load_dataset

ds = load_dataset("criteo/CriteoClickLogs")