---
dataset_info:
  features:
  - name: input_ids
    sequence: int32
  - name: token_type_ids
    sequence: int8
  - name: attention_mask
    sequence: int8
  - name: next_sentence_label
    dtype: int64
  splits:
  - name: train
    num_bytes: 7276477808
    num_examples: 2353324
  - name: validation
    num_bytes: 806789376
    num_examples: 260928
  download_size: 779432338
  dataset_size: 8083267184
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
---