---
dataset_info:
  features:
  - name: documents
    sequence:
      sequence: float16
  splits:
  - name: train
    num_bytes: 8019022928
    num_examples: 171332
  download_size: 5775769873
  dataset_size: 8019022928
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---