---
dataset_info:
  features:
  - name: token_ids
    dtype: uint16
  - name: freq
    dtype: int64
  - name: tokens
    dtype: large_string
  splits:
  - name: train
    num_bytes: 1207207
    num_examples: 50063
  download_size: 1006318
  dataset_size: 1207207
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---