---
dataset_info:
  features:
  - name: input_ids
    sequence: int32
  - name: attention_mask
    sequence: int8
  - name: start_positions
    dtype: int64
  - name: end_positions
    dtype: int64
  - name: sequence_ids
    sequence: int64
  splits:
  - name: train
    num_bytes: 388306912.0
    num_examples: 57544
  - name: test
    num_bytes: 97083476.0
    num_examples: 14387
  download_size: 46592442
  dataset_size: 485390388.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---