# Zero Scrolls dataset

[Zero scrolls](https://www.zero.scrolls-benchmark.com/) includes ten natural language tasks across multiple domains, including summarization, question answering, aggregated sentiment classification and information reordering.

## Hugging Face dataset

The Hugging Face dataset for Zero Scrolls can be found [here](https://huggingface.co/datasets/simonjegou/zero_scroll). To reproduce this dataset, simply run the `create_huggingface_dataset.py` script.

## Evaluation

The answer are not provided in the dataset, you will need to submit your predictions to the [Zero Scrolls](https://www.zero.scrolls-benchmark.com/) website to get the results.