---
dataset_info:
  features:
  - name: subset
    dtype: string
  - name: context
    dtype: string
  - name: qid
    dtype: string
  - name: question
    dtype: string
  - name: detected_answers
    struct:
    - name: char_spans
      list:
      - name: end
        sequence: int64
      - name: start
        sequence: int64
    - name: text
      sequence: string
    - name: token_spans
      list:
      - name: end
        sequence: int64
      - name: start
        sequence: int64
  - name: answers
    sequence: string
  - name: query_embedding
    sequence: float32
  splits:
  - name: train
    num_bytes: 1615608539
    num_examples: 253798
  download_size: 1390791813
  dataset_size: 1615608539
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---