| dataset_info: | |
| features: | |
| - name: context | |
| dtype: string | |
| - name: question | |
| dtype: string | |
| - name: answers | |
| sequence: string | |
| - name: key | |
| dtype: string | |
| - name: labels | |
| list: | |
| - name: end | |
| sequence: int64 | |
| - name: start | |
| sequence: int64 | |
| splits: | |
| - name: train | |
| num_bytes: 100706304 | |
| num_examples: 104071 | |
| - name: validation | |
| num_bytes: 12941478 | |
| num_examples: 12836 | |
| download_size: 61870589 | |
| dataset_size: 113647782 | |
| # Dataset Card for "naturalquestionsshortqa" | |
| Split taken from the MRQA 2019 Shared Task, formatted and filtered for Question Answering. For the original dataset, have a look [here](https://huggingface.co/datasets/mrqa). |