---
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: test
    num_bytes: 1873397
    num_examples: 1503
  download_size: 340899
  dataset_size: 1873397
---
# Dataset Card for "dropqa"

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).