README.md

---
license: mit
dataset_info:
  features:
  - name: id
    dtype: int64
  - name: context
    dtype: string
  - name: question
    dtype: string
  - name: answers
    struct:
    - name: answer_start
      sequence: int64
    - name: text
      sequence: string
  - name: is_impossible
    dtype: bool
  splits:
  - name: train
    num_bytes: 65160295
    num_examples: 2019
  - name: validation
    num_bytes: 39069457
    num_examples: 1212
  download_size: 5496786
  dataset_size: 104229752
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---