---
dataset_info:
  features:
  - name: raw
    dtype: string
  - name: formatted
    dtype: string
  - name: tokenized
    sequence: int64
  - name: attention_mask
    sequence: int8
  splits:
  - name: train
    num_bytes: 2057700892
    num_examples: 24414
  download_size: 214067759
  dataset_size: 2057700892
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---