---
dataset_info:
  features:
  - name: title
    dtype: string
  - name: id
    dtype: string
  - name: context
    dtype: string
  - name: question
    dtype: string
  - name: answers
    struct:
    - name: text
      sequence: string
    - name: answer_start
      sequence: 'null'
  splits:
  - name: train
    num_bytes: 104802744
    num_examples: 77400
  - name: validation
    num_bytes: 11517839
    num_examples: 9534
  download_size: 11491504
  dataset_size: 116320583
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---

Ucinlp's DROP dataset https://huggingface.co/datasets/ucinlp/drop <br>
styled in the format of Rajpurkar's SQuAD data https://huggingface.co/datasets/rajpurkar/squad